DOE FAQ Alert Electronic Newsletter

Issue: Volume 2, Number 5
May 2002
Mark J. Anderson, Stat-Ease, Inc.

Here's another set of frequently asked questions (FAQs) about doing design of experiments (DOE), plus alerts to timely information and free software updates. If you missed previous DOE FAQ Alerts, go to the links below.
Feel free to forward this newsletter to your colleagues. They can subscribe by going to

I offer the following link as an appetizer: It shows the rare alignment of planets that will occur May 5th. Take a look at the website and the sky. Elsewhere at this site you will see how paper plates can be used to explain various aspects of science. Talk about keeping things simple!

Also, here's a follow up to my "appetizer" from last month (see Tom Pyzdek, a reader of this Alert and expert in his own right on SPC and DOE, says, "I'm very interested in the "Global Warming" slant on the ice pack story. Many of your links to news stories on the event play up the GW angle big time. The evidence that global warming exists is quite strong. However, I believe that many around the world employ junk science to advance political agendas. I wrote a couple of columns analyzing hypotheses about the causes of global warming using data that are readily available and simple SPC techniques. The links are: and The articles include discussions of lessons for people applying SPC, or DOE. Hope you find the articles interesting. Tom"

Here's what I cover in the body text of this DOE FAQ Alert (topics that delve into statistical detail are rated "X"):

1. FAQ: Why are no results provided by the ANOVA when all effects are chosen from a two-level factorial design?
2. X-FAQ: A space-filling response surface design selected from a pre-existing set of data points
3. Reader/User feedback: How to do a Plackett-Burman design with less than seven factors
4. Events alert: Stat-Ease is presenting a talk on robust DOE for a Six Sigma session at the Annual Quality Congress in Denver
5. Workshop alert: Upcoming schedule
6. Address alert: New suite number 480

PS. Quotes for the month - Hypotheses by my wife's preschoolers.

1 - FAQ: Why are no results provided by the ANOVA when all effects are chosen from a two-level factorial design?

-----Original Question-----

North Carolina

"Mark, I have a question on the ANOVA table of a factorial design. I have a 2-level factorial. When I choose all factors in the EFFECTS graph including the interaction, then I get no result for "Prob > F" in the ANOVA table. Also the adequate precision is listed as 0.0 and the R-squared is listed as 1.00. Could you please explain why is this happening?"


ANOVA requires a measure of error to determine the F-statistic for overall significance. Two-level factorial designs typically contain no replicates, so if you pick all estimable effect, there will be no measure of error. You should go back to the half-normal plot of effects and click off the smallest ones as error before doing ANOVA. We explain how to do this with Design-Ease® and Design-Expert® in the "Two-Level Factorial Tutorials" section of the User Guides for these software packages. For details on the half-normal plot, click on the following link: (NIST/Sematech "Engineering Statistics Handbook"). For more explanation see Chapter 3 of "DOE Simplified: Practical Tools for Effective Experimentation," by me and my partner Patrick Whitcomb. You can buy this book at

(Learn more about analyzing results from two-level factorial designs by attending the 3-day computer-intensive workshop "Experiment Design Made Easy." For a description, see Link from this page to the course outline and schedule. You can enroll on-line by linking to the Stat-Ease e-commerce page for workshops.)

2 - X-FAQ: Space-filling response surface design selected from a pre-existing set of data points.

-----Original Question-----


"Hi, I have downloaded your trial software, and I am finding your space-filling designs very useful. My problem is that I have 171 data points from which I want to select X number as spread as possible. I don't know X but I know I have an unknown number of near-replicas in some of the 171 points that I don't want included"

Answer (from Pat Whitcomb, consultant and principal of Stat-Ease): "Our distance algorithm is just what you are looking for. It picks points from a candidate set to be spread as far apart as possible. The algorithm starts by picking a vertex. Then each subsequent point picked is the one that has the maximum distance from all previously selected points.

Here is how to select your points:

a. Go to "File" and "New Design".
b. Use the Historical Data option on the "Response Surface" tab to build a design to hold your candidate points:
° Enter the minimum and maximum of each factor as the -1 and +1 levels.
° Enter the number of rows as the size of your candidate set (171).
° Don't enter any response, just click Continue at this stage.
c. Paste in your 171 data points and save the file.
d. Under "Design Tools" choose "Write Candidate File" and save.
e. Select "File" and "New Design". Say "Yes" to "Use previous design info?".
f. Choose the "Distance-Based" option on the "Response Surface" tab and click Continue.
° Click on the "Read list" button and select the candidate file you created containing your 171 points.
° Set the Model, Lack-of-Fit and Replicate points. (If you want a pure distance-based design, set Lack-of-Fit and Replicate points to be zero.) Click Continue and Design-Expert produces the "space-filling" design that you desired.

Remember that with a distance-based design there is no guarantee that the design matrix for a specific model won't be singular. After selecting your design points, click the "Evaluation" node in the software and click on "Results" to see if the model you want can be estimated and, if so, the quality of the design.

(Learn more about developing good RSM designs by attending the "Response Surface Methods for Process Optimization" workshop. For a description, see Link from this page to the course outline and schedule. You can enroll on-line by linking to the Stat-Ease e-commerce page for workshops.)"

3 - Reader/User feedback: How to do a Plackett-Burman design with less than seven factors

Plackett-Burman (P-B) 5-factor Designs

"I am a registered user of Design-Expert and I have a question regarding Plackett-Burman designs. I noticed in the last newsletter* that the issue of how to setup a 8-run P-B with 7 factors was clarified. (Thank you for that by the way!) How can I configure Design-Expert to set up a P-B design with only 4, 5 or 6 factors?"

Answer (from Pat Whitcomb):

"An eight-run Plackett-Burman design is just one of many fractions of the 2^(k-p) two-level fractional factorials. Design-Expert (or Design-Ease) offer no specific P-B designs for less than seven factors, you just use fewer columns in the eight-run P-B and leave the other factors empty ("dummy" factors). A better choice is to use the regular 2^(k-p) fractions. These are minimum aberration designs** (i.e. they have the least confounding). Choosing columns arbitrarily for the P-B designs can (and most often) will lead to poorer designs than those available in the standard two-level design builder."

*(DOE FAQ Alert, Volume 2, Number 2a - March, 2002, "How Design-Expert handles P-B designs." To see this newsletter click on
**(See Montgomery's "Design and Analysis of Experiments," 5th edition, page 326 for details on minimum aberration design. This book can be purchased at

4 - Events alert: Stat-Ease is presenting a talk on robust DOE for a Six Sigma session at the Annual Quality Congress in Denver

I will be in Denver, Colorado for the Annual Quality Congress of the American Society of Quality (ASQ) on May 20 through 22. See for details. Come by and see what Stat-Ease has to offer at booth #725. I hope you can attend the talk written by me and my fellow consultant Shari Kraber on May 20 (Session M205, 4:30 PM). It's entitled "Cost-Effective and Information-Efficient Robust Design for Optimizing Processes and Accomplishing Six Sigma Objectives." The paper we drafted for the conference proceedings can be viewed at

Next month Stat-Ease will be chairing a session on "The Use of DOE in Non-Manufacturing Environments" at the Spring Research Conference of the Quality and Productivity (Q&P) section of the American Statistical Association (ASA) on June 5-7 in Phoenix, Arizona. It will be a great opportunity to hear how DOE and other statistical tools can be used to enhance productivity and improve the quality of products and services. For more details, see (page no longer available). The DOE for non-manufacturing talks will be in Invited Session 9 on Thursday, June 6 from 10:30am to 12:00pm.

Click for a listing of where Stat-Ease consultants will be giving talks and doing DOE demos. We hope to see you sometime in the near future!

5 - Workshop alert: Upcoming schedule

"Experiment Design Made Easy" (EDME) will be presented in San Jose next week, May 7-9. A few seats still remain for this. Otherwise come to Minneapolis for the next presentation, June 4-6, in our new training facility. We return to the West Coast for another EDME in Seattle on July 9-11. For a description of this workshop, see

If you're looking for a quick introduction to DOE, consider our economical, one-day "DOE Simplified" presentation. It will be given in Minneapolis at the Metrodome Holiday Inn on May 30. See for class content.

For those of you who've mastered factorial design and wish to advance your DOE skills, consider coming to one or more of these workshops, all presented in Minneapolis at our new training facility:
- "Mixture Design for Optimal Formulations," May 14-16. If this is too soon, consider the next session on August 13-15. See for details.
- "Robust Design: DOE Tools for Reducing Variability," June 11-13. This is a good class for anyone involved in Six Sigma. Refer to for class content and prerequisites (knowledge of Response Surface Methods required).
- "Response Surface Methods for Process Optimization," July 16-18. See for details and
prerequisites (must be proficient at doing factorial design).

We also present a workshop on other useful statistical tools, called "Statistics for Technical Professionals," on July 23-24 in Minneapolis in Stat-Ease's training room. For a description of this class, see

See for schedule and site information on all Stat-Ease workshops open to the public. To enroll, click the "register online" link at our web site or call Stat-Ease at 1.612.378.9449. If spots remain available, bring along several colleagues and take advantage of quantity discounts in tuition, or consider bringing in an expert from Stat-Ease to teach a private class at your site. Call us to get a quote.

6 - Address alert: New Suite Number 480

Last month Stat-Ease moved up from its former suite on the first floor to a larger one on the 4th floor of the same building. Our new address is:

Stat-Ease, Inc.
2021 E. Hennepin Ave., Suite 480
Minneapolis, MN 55413-2726

Note that the last four digits of our zip code changed along with the suite number. Our telephone, fax and e-mail remain the same. I thought I would be ready to provide photographs of our new computer-intensive training room. However, since we won't hold our first class there until next month, we decided to take our time getting just the right chairs and tables. As you might guess, this involves a certain amount of experimentation. For example, like Goldilocks in the children's story about the three bears, we've all been sitting on a selection of potential chairs for the students. What are the chances we can all agree on one chair?

I hope you learned something from this issue. Address your questions and comments to me at:

Mark J. Anderson, PE, CQE
Principal, Stat-Ease, Inc. (
Minneapolis, Minnesota USA

PS. Quotes for the month - Hypotheses by my wife's preschoolers. She asked her kids, ages 2-4, for their thoughts on what would happen if they did certain experiments with food. Here are a few excerpts:

1. Experiment - What happens when you put an egg in blue water?
a) "It will crack."
b) "It will turn blue." (Is this the scientist in the bunch?)
c) "It will die."
d) "It will float."
e) "It might turn into a human."

2. Experiment - What will happen to cheese at room temperature?
a) "It will turn blue." (So much for the scientist!))
b) "It will get holes in it."
c) "A little mouse will sneak in and eat the cheese."

Trademarks: Design-Ease, Design-Expert and Stat-Ease are registered trademarks of Stat-Ease, Inc.

Acknowledgements to contributors:

- Students of Stat-Ease training and users of Stat-Ease software
- Fellow Stat-Ease consultants Pat Whitcomb and Shari Kraber (see for resumes)
- Statistical advisor to Stat-Ease: Dr. Gary Oehlert (
- Stat-Ease programmers, especially Tryg Helseth (
- Heidi Hansel, Stat-Ease marketing director, and all the remaining staff.

Interested in previous FAQ DOE Alert e-mail newsletters? To view a past issue, choose it below.

#1 - Mar 01, #2 - Apr 01, #3 - May 01, #4 - Jun 01, #5 - Jul 01 , #6 - Aug 01, #7 - Sep 01, #8 - Oct 01, #9 - Nov 01, #10 - Dec 01, #2-1 Jan 02, #2-2 Feb 02, #2-3 Mar 02, #2-4 Apr 02, #2-5 May 02 (see above)

Click here to add your name to the FAQ DOE Alert newsletter list server.

Statistics Made Easy

©2002 Stat-Ease, Inc. All rights reserved.