Issue: Volume 7, Number 9
Date: September 2007
From: Mark J. Anderson, Stat-Ease, Inc., Statistics Made Easy® Blog

Dear Experimenter,

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 the previous DOE FAQ Alert, please click on the links at the bottom of this page. If you have a question that needs answering, click the Search tab and enter the key words. This finds not only answers from previous Alerts, but also other documents posted to the Stat-Ease web site.

Feel free to forward this newsletter to your colleagues. They can subscribe by going to If this newsletter prompts you to ask your own questions about DOE, please address them via mail

For an assortment of appetizers to get this Alert off to a good start, see these new blogs at
—Calculators achieve ‘retro’ status—‘70’s style available again from HP
—Making the most from unhappy events (discusses "the Anna Karenina" (TAK) bias)
—Coming 'round the Smoky Mountains
—The challenge of dealing with statistical anomalies such as bridge collapses

Topics in the body text of this DOE FAQ Alert are headlined below (the "Expert" ones, if any, delve into statistical details).
1. FAQ: Oh, oh! Dropped my flask on the floor! Now what?
2. Expert FAQ: Dr. Frankenstein rebuilds a fractional factorial
3. Webinar Alert: Learn about the factorial planning process
4. Info Alert: Details on DOE for telemarketers, and response surface methods (RSM) for rubber and plastics
5. Job Opening: Stat-Ease seeks a DOE professional!
6. Events Alert: Talk at Euro stats conference
7. Workshop Alert: See when and where to learn about DOE

PS. Quote for the month: Addicted to "DX." (Page through to the end of this e-mail to enjoy the actual quote.)


1. FAQ: Oh, oh! Dropped my flask on the floor! Now what?

-----Original Message-----
From: New York State
"I was wondering what the ramifications were if you botch a run—say you dropped the flask on the floor and were not running any replicates of that flask. Obviously, you can delete that line from your design, but how much will that compromise the rest of the results?"

Answer (from Stat-Ease Consultant Wayne Adams):
"What happens to your design depends on the base design and which point you lose. For fractional factorial designs this will be one of the corners. Often this will result in a two-factor interaction being aliased with everything. The best way to tell is to leave the data missing (if no response could be measured) or ignore the result in question, and then use the design evaluation tool in your Design-Ease® or Design-Expert® software to see what information is lost. Never delete an entire run—instead use the ignore feature in your program's design layout. (I also urge you to make use of the Notes node to record why a particular value is set to ignored status.) Mathematically, any of the following options offered by Stat-Ease software will produce the same model:
—Ignore (via right-click to Cell Status and change from Normal),
—Leaving the cell blank (missing), and
—Delete row (via right-click on button to left)
However, it's always best to preserve the factor levels and measured response value (if any) by setting individual cells or the entire run to be ignored (shown lined out by 'strike-through' in the software's display of the design layout)."

Further comments:
I cannot resist telling this related anecdote. Years ago I worked as a process development engineer in a specialty-chemical research center. I had the great fortune of working in a department that encouraged use of statistically designed experiments and provided on-the-job training on these powerful tools. However, some of the chemists lacked enthusiasm for DOE. I arm-twisted one into running a two-level factorial on processing conditions at the bench prior to me scaling it up for pilot-plant studies. Analyzing his data, I found an extreme outlier, which upon inspection of the lab notebook, was described thusly: "I initiated the reaction successfully, but the alarm went off for a fire drill and I accidentally knocked the beaker to the floor. Half an hour later I picked out the glass, transferred the viscous mass to another beaker and resumed the reaction." I deemed this a special cause and replaced the outlier (exceedingly low for yield!) with an imputed value—the old-fashioned approach for dealing with missing data. Thank goodness for dedicated DOE software and computers capable of dealing better with incomplete test matrices.

(Learn more about dealing with apparent outliers by attending the three-day computer-intensive workshop "Experiment Design Made Easy." See for a description of this class and then link from this page to the course outline and schedule. Then, if you like, enroll online.)


2. Expert-FAQ: Dr. Frankenstein rebuilds a fractional factorial

-----Original Message-----
From: "Your Friendly, Neighborhood Statistician" Arved Harding"
I was doing some interesting design stuff today with version 7.1 of Design-Expert (DX). I felt like Dr. Frankenstein after seeing the new feature that changes colors on the factorial design builder when you change the number of blocks in the factorial. [For example, adding blocks may degrade the resolution from
'green' resolution V to 'yellow' resolution IV.] I was playing with our old buddy the 2^5-1 design. The default generators in DX are I = ABCE and Block=ABD. This causes you to have 3 two-factor interactions (2FI) confounded with other 2FIs. I found an NBS Standard from 1957 that used as the default I=ABCDE and Block =AB. So you only lose one interaction this way. I realize that confounding interactions with one another is not as serious as confounding interactions with Block, however you only lose one interaction if you chose this design. If you have two factors that can't logically interact, then this may be a better design."

Answer (from Stat-Ease Consultant Pat Whitcomb):

"Dear Dr. Frankenstein, Thanks for the feedback and suggestions; glad you're enjoying yourself! We once used a 2FI for the block generator; then minimum aberration came along. In minimum aberration the lowest order terms are the most important and get aliased with the higher order terms. The order is that ME is a
1, a 2FI is 2, a 3FI is 3, etc. with blocks being a 1.5. Therefore having a block term generated by a 3FI (the new scheme) rather than a 2FI (the old scheme) is lower aberration. Having thought about this for a bit it makes sense to me. (My first reaction was identical to yours.) Let's say AB is used to generate blocks; and there is a significant AB effect. The block effect now has a random component (whatever was blocked on) and a fixed component (AB). If the AB effect is interpreted as a blocking problem, time is spent trying to determine why the blocks (machines, lots, operators or the like) are different and the fact there is a significant 2FI is missed. Using ABD to generate blocks gives some assurance the block mean square (MS) is due to what was blocked on and isolates the 2FIs. Hopefully, you get some idea about the existence of 2FIs, even if they can't be positively identified. So, in the end, I bought into the definition for minimum aberration."


3. Upcoming Webinar: Learn about the factorial planning process

You are invited to attend a free web conference by Stat-Ease on "A Factorial Design Planning Process" at 8 AM and 11 AM USA Central Daylight Time (CDT) on Wednesday, September 19.* It will be based on a talk presented at last year's Fall Technical Conference (FTC), which is jointly sponsored by American Society of Quality (ASQ) Chemical & Process Industries Division, the ASQ Statistics Division, and the American Statistical Association (ASA) Sections on Physical and Engineering Sciences and Quality and Productivity. Stat-Ease Consultant Shari Kraber, who co-presented with Pat Whitcomb, reported afterwards that: "The Fall Technical Conference was attended by about 200 people; a small, but highly-influential group of industrial and academic statisticians. We presented the talk, which provides a structure for planning designs and incorporates our new up-front power calculations. Lots of positive comments were given afterwards. Another speaker was so intrigued by our unique minimum resolution V (MR5) designs that he added a citation to his own presentation, triggering an interested reaction out of that audience as well."

Attendance may be limited for one or both of the two one-hour long webinar sessions on "A Factorial Design Planning Process." Contact me via to sign up. If you can be accommodated, I will send you the link for the WebConnect and dial-in for ConferenceNow voice via telephone. Toll-free access extends world-wide, but not to all countries. Details will be provided as needed.

Here is the talk's description provided by Shari and Pat to FTC attendees:
Purpose: To present a specific process to organize and plan a designed experiment, taking into consideration the power of the design relative to the size of the effects.
Abstract: Newcomers to factorial design find it difficult to choose appropriate designs with adequate power. In this presentation, we introduce a clear process to determine the best design that fits the problem. A discussion of statistical power will show attendees how the size of the effect relative to the noise is a critical criterion in design selection. We will also discuss how to choose the ranges for the input factors, the importance of evaluating aliases, and checking runs for safety. Various case studies will illustrate the importance of using this process to avoid DOE failure due to an incorrect design choice. Attendees will take away a clear strategy for determining which factorial design is appropriate for their data analysis needs.

*Argghh! Mark, here, being a bit of a buccaneer throwing a marlin spike into the works of this here webinar: It sails on "Talk Like a Pirate Day" (, so if my Stat-Ease mateys say such things as "Avast, ye stat-lubbers, keep a sharp eye for yonder optimum—thar be the treasure!" do not jump overboard! ;)


4. Info Alert: Details on DOE for telemarketers, and response surface methods (RSM) for rubber and plastics

Two new articles on statistical tools for design of experiments are now available for your educational purposes.

Bob Gahagan, an independent consultant operating out of Columbia, South Carolina, gave Stat-Ease permission to post an inspiring article on DOE for telemarketers that you will find at

In the Technical Notebook of their August 20th issue, Rubber And Plastics News ( published an article by me titled "Response Surface Methods for Peak Process Performance." See my text at
This primer on RSM is the third in a series on DOE dating back a decade,* which I wrote (with Pat Whitcomb's help) by invitation from editor Harold Herzlich. He says "The object is to help rubber-industry technical people understand and more fully utilize the more efficient design of experiment (DOE) statistical
techniques that are developing. I have found many rubber industry not aware of the way new software can take the pain out of the process while helping design and evaluate multi-factor research, development or industrial studies."

*Others are:
—"Breakthrough improvements with experiment design" published in June 1997, see for the original manuscript
—"Mixture DOE uncovers formulations quicker" published on October 21, 2002, see

PS. In last month's DOE FAQ Alert I announced publication of the second edition of "DOE Simplified." The link I provided went directly to the publisher—Productivity Press of New York City. I am pleased to see that it has now been picked up by ASQ's Quality Press and posted at their web site. Go to for details on this fun, but very informative primer on DOE, and link from there to purchase it online.


5. Job Opening: Stat-Ease seeks a DOE professional!

Stat-Ease, Inc., a Minneapolis-based DOE software, training and consulting company, has a permanent, full-time opportunity for an energetic person to join our team. The position is a combination of technical development and teaching. Job responsibilities include teaching design of experiments workshops, providing statistical support to clients, defining test cases for statistical software development, researching statistical methods for future software implementation, conveying statistical concepts to programmers, writing/editing technical materials, and other duties based on experience. This is a full-time, permanent
position in Minneapolis, MN with approximately 25% travel.

Minimum qualifications:
—Chemical or Biological or related Engineering degree with 5-10 years experience
—Strong math skills
—Aptitude and desire to learn more about how statistics can improve products and processes
—Proven ability to turn concepts into solutions
—Hands-on DOE experience, mixture design desirable
—Teaching experience and strong ability to communicate with technical professionals
—Excellent English skills—verbal and written
—US citizenship required

Resumes should be sent to


6. Workshop alert: See when and where to learn about DOE

Seats are filling fast for the following DOE classes:

Experiment Design Made Easy (EDME) (Detailed at
—September 18-20 (Philadelphia, PA)
—October 9-11 (Minneapolis, MN)

Mixture Design for Optimal Formulations (MIX) (
—October 23-25 (Minneapolis)

Response Surface Methods for Process Optimization (RSM) (
—September 25-27 (Minneapolis) **SOLD OUT**
—November 13-15 (Minneapolis) — added to accommodate overflow from September class — seats available!

DOE for DFSS: Variation by Design (DDFSS) (
—November 7-8 (Minneapolis)

See for complete schedule and site information on all Stat-Ease workshops open to the public. To enroll, click the "register online" link on our web site or call Elicia at 612.746.2038. 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 a private workshop, which is most convenient and effective for your staff. For a quote, e-mail


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



Mark J. Anderson, PE, CQE
Principal, Stat-Ease, Inc. (
2021 East Hennepin Avenue, Suite 480
Minneapolis, Minnesota 55413 USA

PS. Quotes for the month—Addicted to "DX":
(Note: A technical manager at General Electric said his employees are "addicted" to Design-Expert ("DX"), thus precipitating a renewal of their annual network license to satisfy their addiction. Our resident punster, Tryg Helseth, was moved by this news to adapt the lyrics to a popular song. —Mark)

Addicted to DX
by Tryg Helseth (with apologies to Robert Palmer*)

"Your computer's on, you're data's in
Your mind is lost amidst the din
Your palms sweat, your keyboard shakes
Another transform is what it takes
You can't sleep, you can't eat
You have runs you must repeat
You squeeze the mouse, it starts to squeak
A new analyses will look less bleak
Whoa, you like to think that you're immune to the stats, oh yeah
It's closer to the truth to say you need your next fix
You know you're gonna have to face it, you're addicted to Dee Ex

* Author of lyrics at

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
—Stat-Ease consultants Pat Whitcomb, Shari Kraber and Wayne Adams (see for resumes)
—Statistical advisor to Stat-Ease: Dr. Gary Oehlert (
—Stat-Ease programmers, especially Tryg Helseth and Neal Vaughn (
—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, #2-6 Jun 02, #2-7 Jul 02, #2-8 Aug 02, #2-9 Sep 02, #2-10 Oct 02, #2-11 Nov 02, #2-12 Dec 02, #3-1 Jan 03, #3-2 Feb 03, #3-3 Mar 03, #3-4 Apr 03, #3-5 May 03, #3-6 Jun 03, #3-7 Jul 03, #3-8 Aug 03, #3-9 Sep 03 #3-10 Oct 03, #3-11 Nov 03, #3-12 Dec 03, #4-1 Jan 04, #4-2 Feb 04, #4-3 Mar 04, #4-4 Apr 04, #4-5 May 04, #4-6 Jun 04, #4-7 Jul 04, #4-8 Aug 04, #4-9 Sep 04, #4-10 Oct 04, #4-11 Nov 04, #4-12 Dec 04, #5-1 Jan 05, #5-2 Feb 05, #5-3 Mar 05, #5-4 Apr 05, #5-5 May 05, #5-6 Jun 05, #5-7 Jul 05, #5-8 Aug 05, #5-9 Sep 05, #5-10 Oct 05, #5-11 Nov 05, #5-12 Dec 05, #6-01 Jan 06, #6-02 Feb 06, #6-03 Mar 06, #6-4 Apr 06, #6-5 May 06, #6-6 Jun 06, #6-7 Jul 06, #6-8 Aug 06, #6-9 Sep 06, #6-10 Oct 06, #6-11 Nov 06, #6-12 Dec 06, #7-1 Jan 07, #7-2 Feb 07, #7-3 Mar 07, #7-4 Apr 07, #7-5 May 07, #7-6 Jun 07, #7-7 Jul 07, #7-8 Aug 07, #7-9 Sep 07 (see above)

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

Statistics Made Easy®

DOE FAQ Alert ©2007 Stat-Ease, Inc.
All rights reserved.


Software      Training      Consulting      Publications      Order Online      Contact Us       Search

Stat-Ease, Inc.
2021 E. Hennepin Avenue, Ste 480
Minneapolis, MN 55413-2726
p: 612.378.9449, f: 612.378.2152