Example of Analyze Definitive Screening Design

Engineers are developing a new ultrasonic cleaner. The engineers use a screening design to determine which potential factors affect the output power for the cleaner.

One of the engineers analyzes a definitive screening design to determine which of 7 factors have the greatest effect on the power output. Power output needs to be high enough to clean adequately. At the same time, power output needs to be low enough to clean without damaging the items.

  1. Open the sample data, ultrasonic_cleaner.MTW.
  2. Choose Stat > DOE > Screening > Analyze Screening Design.
  3. In Responses, enter Power.
  4. Click Terms.
  5. In Include the following terms, choose Linear. Click OK.
  6. Click Graphs.
  7. Under Residual Plots, select Four in one.
  8. Check Residuals versus variables. Enter Quiet and Sweep.
  9. Click OK in each dialog box.

Interpret the results

In the Pareto chart, the engineer sees that the largest main effects are for Train (A) and Quiet (D). From the screening experiment, the engineer concludes that these two factors deserve the most consideration for further analysis.

The residual versus fits plot shows a U-shaped curve. This pattern is an indication that the model could be missing square terms or interactions. The plots of the residuals against Quiet and Sweep show curves also. In the exploration of potential models, the engineer decides to consider the square terms for these factors.

Screening design model: Power versus Train, Degas, Burst, Quiet, Center, ...

Analysis of Variance Source DF Adj SS Adj MS F-Value P-Value Model 7 61803.7 8829.1 3.40 0.046 Linear 7 61803.7 8829.1 3.40 0.046 Train 1 38451.1 38451.1 14.80 0.004 Degas 1 11.2 11.2 0.00 0.949 Burst 1 1040.6 1040.6 0.40 0.542 Quiet 1 21938.4 21938.4 8.45 0.017 Center 1 77.8 77.8 0.03 0.866 Bandwidth 1 171.5 171.5 0.07 0.803 Sweep 1 113.1 113.1 0.04 0.839 Error 9 23375.4 2597.3 Total 16 85179.2
Model Summary S R-sq R-sq(adj) R-sq(pred) 50.9634 72.56% 51.21% 4.50%
Coded Coefficients Term Coef SE Coef T-Value P-Value VIF Constant 657.8 12.4 53.22 0.000 Train 52.4 13.6 3.85 0.004 1.00 Degas 0.9 13.6 0.07 0.949 1.00 Burst 8.6 13.6 0.63 0.542 1.00 Quiet -39.6 13.6 -2.91 0.017 1.00 Center -2.4 13.6 -0.17 0.866 1.00 Bandwidth 3.5 13.6 0.26 0.803 1.00 Sweep 2.8 13.6 0.21 0.839 1.00
Regression Equation in Uncoded Units Power = 626.5 + 116.5 Train + 2.0 Degas + 1.92 Burst - 8.80 Quiet - 0.47 Center + 0.70 Bandwidth + 0.57 Sweep
Alias Structure (up to order 2) Factor Name A Train B Degas C Burst D Quiet E Center F Bandwidth G Sweep
Aliases I + 0.82 AA + 0.82 BB + 0.82 CC + 0.82 DD + 0.82 EE + 0.82 FF + 0.82 GG A B C D E F G
By using this site you agree to the use of cookies for analytics and personalized content.  Read our policy