Dear Experimenter,

Here’s a fresh set of answers to frequently asked questions (FAQs) about design of experiments (DOE); plus, timely alerts for events, publications, and software updates. Check it out!

Please let me know what you learned from this issue: I’d really appreciate hearing from you! Address your questions and comments to me at

Please do not send me requests to subscribe or unsubscribe, follow the instructions at the end of this message.

Mark J. Anderson, PE, CQE
Engineering Consultant, Stat-Ease, Inc.

PS Quote for the month:

Secret of success for Amazon revealed by founder Jeff Bezos.

(Page down to the end of this e-zine to enjoy the actual quote.)

Vol. 21, No. 1 - Jan/Feb 2021

Software Alert
v13 of Design-Expert Released!

Identifying the Vital Few Factors from a Screening Design

Webinar Alert 
Free Educational DOE Webinars

Info Alert
DX13 Poisson Regression Maximizes Popcorn Output

Workshop Alert
Enroll Before Spring Classes Fill

StatsMadeEasy Blog
My wry look at all things statistical and/or scientific with an engineering perspective.
Also, see the Stat-Ease blog for tips on making DOE easy. For example, a recent posting provides an answer to “Christmas trees on my effects plot?”. Take a look!
Software Alert
Version 13 of Design-Expert® software released!
The newly released version 13 of Design-Expert software (DX13) provides a substantial step up on ease of use and statistical power for design and analysis of experiments. Here are my favorite new tools out of the many upgrades:
  • Design space augmentation—Refocus an existing experiment to the most promising region by expanding, shrinking or moving your design space, whether it be mixture or process space.
  • Poisson regression—Fit counts most precisely, for example, unpopped kernels from microwave popcorn as reported in this issue of the DOE FAQ Alert.
  • Multiple analyses per individual response—Easily model any given response various ways to readily compare them.
  • Rounding for factor or component settings—Incredibly convenient for most designs but especially so for custom-optimal ones that generate levels to many decimal places. Amazingly, it works just as well for mixture as for process studies.
  • Import data wizard—With a press of a few buttons, simply paste in your valuable existing records from a spreadsheet or another statistical program. Then, with a bit of cleanup and identification of inputs versus outputs, you are good to go with Design-Expert’s advanced tools for design evaluation, modeling, visualization and optimization.
See more about DX13 and its amazing features at its home page. From there you can view license options, go for a fully functional, free trial or request a demo. If you own a subscription or network license, then upgrade it to v13 via this site.

Make the most from every experiment with version 13 of Design-Expert!

PS To see all the details on revisions to Design-Expert, navigate via Help, Contents to What’s New (press the “+” under Full Changelog for a complete list of version changes). If you want to receive notice when an update becomes available, go to Edit on the main menu of your program, select Preferences and, within the General tab (the gear icon at the right), turn on (if not already on by default) the “Check for updates on program start” option.
Identifying the Vital Few Factors from a Screening Design

Original question from an Expertise Lead:
“We used Design-Expert to screen 10 factors via a standard, 32-run, fractional, two-level design. I attended your webinar on How to Detect and Overcome Bad Data and, with the know-how provided, managed to get a good model by transforming the response per the Box-Cox plot, as well as ignoring one outlier. The experiment revealed several two-factor interactions that create a significant impact our product’s functionality. I would like to follow up on the vital few factors with a characterization design. Please advise how to proceed.”

Ideally, as I presented in my talk, you can assign a special cause before ignoring a statistical outlier. However, in some cases, such as yours, this becomes moot: It makes no difference to the outcome whether you keep this result or discard it. Therefore, I do not dispute you taking out the outlier. Furthermore, I applaud you using the square root transformation: Good catch!

However, I am concerned that your two-factor interactions (2FIs) are aliased, this being a yellow (proceed with caution!), Resolution IV, screening design. In this case, DX simply labels 2FIs with the one being lowest in alphabetic order. You can easily switch the labeling on half-normal effects plots by right-clicking the symbol. (See, for example, the Biker-Foldover case available via Help, Tutorials from within DX.) After exploring the alternatives, I came up with an alternative view based on heredity—a general rule that strong interactions tend to have both parents active. My model includes only 4 of the original 10 factors, whereas yours requires 6 factors. You, being the subject matter expert, must judge by viewing the alternative interaction plots which model to choose, but moving ahead to the next phase of characterization will require fewer runs with my simpler and more sensible model.”
*PS Professor Erik Vantahalo of Lulea University provides a great 7-minute lecture on effect sparsity, hierarchy and heredity principles in this YouTube video. Keep in mind that opinions vary on adherence to these ‘rules’.
(Learn more about process screening by attending the next distance-learning presentation of Modern DOE for Process Optimization.)
Check out these highly educational presentations from our world-class DOE experts:
  • February 11—“Keys to Analyzing a Response Surface Design” by Shari Kraber
  • February 17—“Cutting Edge Tools Unveiled in Design-Expert Version 13” by Martin Bezener
  • February 24—“New-User Intro to Design-Expert Software” by Richard Williams
  • March 10—“Making the Most from Measuring Counts” by me.
See the times, descriptions and registration links for these upcoming live webinars here. Sign up now to advance your DOE know-how!
PS Do you need a speaker on DOE for a learning session within your company or professional society at regional, national, or international levels? If so, please get back to me. – Mark
DX13 Poisson Regression Maximizes Popcorn Output
Energized by new tools in version 13 of Design-Expert (DX13) for modeling counts, Engineering Consultant Mark Anderson tests a cellphone app against built-in timing on his microwave for minimizing unpopped kernels (UPK). DX13 paves the way to nearly perfect popcorn via its precise analysis via Poisson regression.

To see the surprising, but very satisfying, outcome, click here.
Sharpen up on DOE—Enroll before spring classes fill

You can do no better for quickly advancing your DOE skills than attending a Stat-Ease workshop. Our expert instructors provide you with a lively and extremely informative series of lectures interspersed by valuable hands-on exercises. Enroll early to ensure your spot! See this web page for the complete schedule of upcoming Stat-Ease distance-learning courses. To enroll in the workshop that suits you best, click Register on that webpage, or click here to contact us.
PS If you lead a group of 6 or more colleagues, save money and customize content via a private workshop. For a quote, please contact us
“Our success at Amazon is a function of how many experiments we do per year, per month, per week, per day.”
—Jeff Bezos, who, according to this Harvard Business Review report on how to do smart business experiments, fired a group of web designers who changed an Amazon page without first doing testing.
Stat-Ease, Design-Expert and Statistics Made Easy are registered trademarks of Stat-Ease, Inc.

Circulation: Over 4000 worldwide
Copyright © *|CURRENT_YEAR|* *|LIST:COMPANY|*, All rights reserved.

Our mailing address is:

Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list.