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!

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


Mark J. Anderson, PE, CQE
Principal, Stat-Ease, Inc.

P.S. Please do not send me requests to subscribe or unsubscribe. Follow the instructions at the end of this email.

Software Alert
Version 12.0.7 of Design-Expert® software released

FAQ Alert
Can Design-Expert estimate the block effect when no runs are replicated across blocks?

Event Alert
8th European DOE Conference coming to Groningen, Netherlands in June 2020

Webinar Alert

Free seminars that span the range from Design-Expert beginners to advanced tools

Workshop Alert
Springing up in Austin at ideal time for flourishing flowers and DOE know-how

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 Awesome Week of Data Analytics!”. Check it out!
Version 12.0.7 of Design-Expert® software released
Current Design-Expert v12 (DX12) users will do well by updating to 12.0.7 via the Help menu from within your installed program. View the change log for details on the changes—mainly maintenance items. 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 default General tab, turn on (if not already on by default) the “Check for updates on program start” option. (This alert feature is not available in network versions.)

If you remain mired in an obsolete version or are not yet a user of Design-Expert, click here for the highlights on what’s new in DX12 (plus more details on added features via this link. Then, assuming you like what you see, follow the FREE TRIAL link for a 30-day tryout or buy it directly, taking advantage of upgrade pricing if eligible.

You will do well by gaining the leading edge on DOE capability by upgrading your software to DX 12.0.7!
Can Design-Expert estimate the block effect when no runs are replicated across blocks?
Original question from a Senior Biostatistician:
“We want to use Design-Expert’s (DX) design tools to optimally augment 17 runs of historical (happenstance) process data with a block of controlled runs that will provide a statistically sound predictive model. For estimating the block effect, I expected one or more of new run being exactly at the same settings the historical data. However, DX laid out a series of all new settings. How is this possible?”

Standard designs such as central composites generally do duplicate runs in each block, typically the center point. However, although this makes it easy to see block effects at a glance, they can, in any case—even in completely unreplicated designs, be estimated mathematically by the least-squares fitting to the blocking term. See this detailing in the NIST Engineering Statistics Handbook. Nevertheless, as a 'control', it would be prudent for you to manually insert a run matching one in your first block into the second block. -Mark

PS. My answer not being completely satisfying, I asked University of Minnesota Statistics Professor Gary Oelhert, author of A First Course in Design and Analysis of Experiments, for this further explanation (thank you Gary!):

“You are correct that the block effect can be estimated even with no design points in common across blocks. The main issue is that we are assuming that the same linear model with the same model coefficients determines the structure of the mean across all blocks, even if the blocks do not share common design points. The only exception to this is that the intercept (or overall mean in some parameterizations) is allowed to vary from block to block. This allowable difference from block to block represents the block effect; some blocks read high, some blocks read low.

"Roughly speaking, fit the same model to all of the data simultaneously. Then add indicator (dummy) variables for blocks. That is the model for blocked data. (This assumes that blocks are fixed. For random blocks, as with inter-block recovery, you change the covariance among the responses, with responses within blocks having positive covariance.)

"Having the same design point in different blocks certainly can make estimation of the block effect more obvious, but this kind of overlap is not required. Having repeated design points within a block allows estimation of pure error, which is useful in assessing lack of fit.”

(Learn more about design augmentation and blocking by attending the three-day computer-intensive workshop on Modern DOE for Process Optimization. Click the title for a description of this class and registration details.) 
8th European DOE Conference coming to Groningen, Netherlands in June 2020—Register now before it sells out and consider giving a talk (last call for papers!)
To learn and, perhaps speak on, how to “Make the Most from Every Experiment” reserve June 18-19 on your calendar for the 8th European DOE Conference in Groningen, Netherlands. Stat-Ease reseller Science Plus Group reserved the exclusive Het Kasteel (The Castle) in the City Center for our gathering of Design-Expert software users and experts in DOE. Come for the meetings and leave time before and/or after to tour this historic European region. For more information on the 8th European DOE Conference and the first call for papers, see this meeting site. We hope to see you in the Netherlands next summer.

Click here for these and other upcoming appearances by Stat-Ease professionals

P.S. Do you need a speaker on DOE for a learning session within your company or technical society at regional, national, or even international levels? If so, contact me. It may not cost you anything if Stat-Ease has a consultant close by, or if a web conference will be suitable. However, for presentations involving travel, we appreciate reimbursement for travel expenses. In any case, it never hurts to ask Stat-Ease for a speaker on this topic.
Free seminars that span the range from Design-Expert beginners to advanced tools—something for everyone, from new users to expert DOE practitioners
Go to the Stat-Ease webinar site to enroll in the presentation that best fits your knowledge level for DOE and Design-Expert:
  • “New User Intro to Design-Expert® Software” by Stat-Ease Trainer Richard Williams on February 12
  • “KCV Models for Combined Mixture Process Designs” by Stat-Ease Consultant Pat Whitcomb on March 4 and 5—three sessions.
Refer to our webinar site for specific times. We hope you will join us for one or more of these enlightening seminars. 
Springing up in Austin at ideal time for flourishing flowers and DOE know-how
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 discussion of topics interspersed by valuable hands-on exercises. Enroll at least 6 weeks prior to the date so your place can be assured—plus get a 10% “early-bird” discount. See our complete schedule on all Stat-Ease workshops open to the public. To enroll, scroll down to the workshop of your choice and click on it, or email our Lead Client Specialist Rachel Poleke at 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.*

*Once you achieve a critical mass of about 6 students, it becomes very economical to sponsor an on-site workshop, which is most convenient and effective for your staff. For a quote, email 
Quote of the Month
“1.  If something can go wrong in conducting an experiment, it will.
2.  The probability of successfully completing an experiment is inversely proportional to the number of runs.
3.  Never let one person design and conduct an experiment alone, particularly if that person is a subject-matter expert in the field of study.
4.  All experiments are designed experiments; some of them are designed well, and some of them are designed really badly. The badly designed ones often tell you nothing.
5.  About 80 percent of your success in conducting a designed experiment results directly from how well you do the pre-experimental planning.
6.  It is impossible to overestimate the logistical complexities associated with running an experiment in a 'complex' setting, such as a factory or plant.”

Doug Montgomery, supplemental text material for his 6th edition of Design and Analysis of Experiments.
statistics made easy
Make the most from every experiment

Stat-Ease, Design-Expert and Statistics Made Easy are registered trademarks of Stat-Ease, Inc.
Acknowledgements to contributors:
Students of Stat-Ease training and users of Design-Expert software
Stat-Ease consultants: Pat Whitcomb, Martin Bezener, and Shari Kraber
Stat-Ease programmers: Hank Anderson, Joe Carriere, and Mike Brownson
Stat-Ease business staff: Cathy Hickman, Greg Campbell, and Rachel Poleke
—who provide such supreme support!

Circulation: Over 7500 worldwide
To subscribe, go to
To unsubscribe, go to the footer of this email for email preferences
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.