Creating a Signal-to-Noise Ratio Response
You can define the following response functions in ADX:
- Mean: the response mean at each control factor design point over the noise array
- Standard Deviation: the response standard deviation at each control factor point
- Variance: the response variance at each control factor point
- Difference from Target: the difference between the average response and a user-specified target at each control factor point
- S/N: Smaller the Better: The "smaller-the-better" signal-to-noise ratio, which increases as the response mean decreases and variation around the mean decreases
- S/N: Nominal the Best: The "nominal-the-best" signal-to-noise ratio, which increases as the response mean gets closer to a user-specified target and variation around the mean decreases
- S/N: Larger the Better: The "larger-the-better" signal-to-noise ratio, which increases as the response mean increases and variation around the mean decreases.
- S/N: Signed Target: The "signed target" signal-to-noise ratio, which is a -10 times the logarithm of the response variance at each inner array design point. This increases as the variation around the mean decreases.
If you click the Signal-to-Noise button in the Outer Array Design Specification window, an ADX message appears that explains how to define the signal-to-noise ratio.
To define the signal-to-noise, follow these steps:
- Go to the main design window.
- Click Define Variables.
- Click the Response tab.
- Click FORCE to select the FORCE row.
- Click S/N Functions.
- Select S/N: Larger the Better since a higher pull-off force means a better quality of nylon tubing; also select Mean since Taguchi (1978) suggested that you should always analyze the mean.
- Click OK to define the signal-to-noise ratio.
- Click OK in the Define Variables window to return to the Mixed Level Design window.
At this point, you have completed all the steps required to create the product design for the pull-off force experiment. To proceed, close this design and open the sample design named pull-off force, which is identical to the design you just created and contains the response values.
Copyright © 2008 by SAS Institute Inc., Cary, NC, USA. All rights reserved.