Issue: Volume 2, Number 9
Date: September 2002
From: 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, click on the links below. Feel free to forward this newsletter to your colleagues. They can subscribe by going to http://www.statease.com/doealertreg.html.
I offer the following link to the
National Inventor's Hall of Fame as an appetizer: http://www.invent.org/index.asp.
One of the 2002 inductees in the ceremony scheduled for the 21st of this month
is Jim Wynne. As a boy he wanted to be like Doctor Zarkov, a physicist who assisted
Flash Gordon in the old movie serials Wynne saw on TV. But his real fascination
was death rays, which as he says, were "a light beam [that] would come
out of a machine and blow things up." (See more on death rays at
http://www.dvdcult.com/rev_FG_DRFM.htm. this link no longer exists)
The old saying that the "difference between men and boys is the cost of
their toys" certainly held true for Wynne, who will be honored for helping
to invent not death rays (thank goodness), but LASIK eye surgery. (An aside
- see why you need sharper eyes for computer devices of the future at http://www.nature.com/nsu/020819/020819-5.html.)
However, invention is not only a 'guy thing.' For example, some years ago the
Hall of Fame honored Nobel prize-winner Gertrude Belle Elion as its first woman
inductee. Elion invented drugs to fight leukemia and facilitate kidney transplants
while working at Burroughs-Wellcome (now GlaxoSmithKline, aka "GSK").
For more on Elion, see http://www.achievement.org/autodoc/page/eli0pro-1.
For the full report on the Inventors Hall of Fame class of 2002, see http://www.americanheritage.com/it/2002/02/hofr.shtml (Update 3/07: Link no longer active).
Here's what I cover in the body text of this DOE FAQ Alert (topics that delve into statistical detail are rated "X"):
1. Stat-Teaser alert: Preview
our upcoming newsletter featuring a science project on flying disks, plus learn
how to avoid DOE disaster
2. FAQ: When to designate a factor as numeric versus categorical
3. X-FAQ: Augmenting a fractional two-level factorial design with the runs that were not performed originally
4. User feedback: Overcoming reluctance by Design-Expert® software to extrapolate
5. Simulation alert: Brew better beer by doing DOE
6. Workshop alert: "Experiment Design Made Easy" will be in Philadelphia (plus a link to other locations and/or other DOE workshops)
7. Events alert: A heads-up on DOE talks in Akron (9/11!), Chicago, and the Philadelphia area (Valley Forge)
8. Reader contribution for the anniversary of the September 11th tragedy: The story of a scientist from India who made a new life in the USA
PS. Quote for the month - David Letterman on USA Today stats
1 - Stat-Teaser alert: Preview
our upcoming newsletter featuring a science project on flying disks, plus learn
how to avoid DOE disaster
Many of you will get a printed copy of the latest Stat-Teaser, so this will be a 'sneak preview.' Others, particularly those who reside outside of North America, get only the electronic copy, which can now be viewed at http://www.statease.com/newsltr.html. The feature article, "Sixth-Graders Experiment with Flying Disks" details a fun study done by my youngest daughter for extra credit at school. Coincidentally, the inventor who added a key aerodynamic feature to the "Frisbee" brand disk died recently. Believe it or not, he asked that his ashes to be shaped into Frisbees which may be put up for sale! See for yourself at http://news.bbc.co.uk/1/hi/world/americas/2193093.stm.
The other story in the Stat-Teaser, "How to Avoid DOE Disaster," is an offshoot from a round-table discussion chaired by our consultant Shari Kraber at the recent Joint Statistical Meetings (JSM) in New York City. She and other experts on DOE talked about the perils and pitfalls that often trap the unwary experimenter doing DOE for the first time. Read what Shari wrote on this if you want to improve your chances for a successful experiment.
2 - FAQ: When to designate a factor
as numeric versus categorical
From: San Francisco
"I am putting a 5-factor, 2-level factorial together. One of my factors is bead size. We have beads which are sieved into two different size ranges - let's call them small versus large. I want to run one level with 100% of the large beads and the other with 10% of the small beads mixed in with 90% of the large beads. My question is: Can I call the factor numeric even though they would be percentages (100 versus 90/10), or is there a better way to do this?"
Answer (from Shari Kraber,
Stat-Ease statistical consultant):
"Since you are doing a 2-level factorial design, it would not matter if the factor was designated as either numeric or categorical. The analysis result will simply show if there is a change in the response as you move from the low setting to the high setting. Quantifying them as 90% and 100% is fine. Just use something that makes sense to you so you can properly interpret the results."
(Learn more about fractional two-level factorial designs by attending the 3-day computer-intensive workshop "Experiment Design Made Easy." See http://www.statease.com/clasedme.html for a description. Link from this page to the course outline and schedule. Then, if you like, enroll online.)
3 - X-FAQ: Augmenting a fractional
two-level factorial design with the runs that were not performed originally
From: San Diego
"As a graduate of one of your incredible DOE courses, I sure hope I can ask this question correctly. When setting up a 5-factor, 1/2 fractional (and 1 replicate for discussion purposes), the data matrix will show 16 runs, treatments, or combinations. Question: If, for some good reason, I wanted to list the opposite or unused 16 treatments, could I do this automatically with Design-Expert version 6 software?"
First of all - flattery will get you everywhere! Secondly, you ask a good question, one that I've never heard before. My first inclination, which proved successful, was to bring up the 2^5-1 design on DX6 and then go to Design Tools, Augment Design and select the Fold over option. This did the job, but it's not the answer for other designs, particularly those that are Resolution IV. I asked my partner Pat Whitcomb, head statistical consultant at Stat-Ease, to come with a more comprehensive answer. Here is what Pat advises:
"To generate the alternate fraction:
1. First go to Design Evaluation and look at the defining contrast.
2. For a half fraction there is only one word in the defining contrast.
3. Using any letter in that word to do a single factor foldover generates the alternate fraction.
For more highly fractionated designs the same idea holds. Look at the defining contrast and try to pick letters that appear in odd combinations. The goal is to generate the negative words in the folded fraction to cancel words in the original fraction."
4 - User FAQ: Overcoming reluctance by Design-Expert software to extrapolate
"I need to extrapolate, to see which level to set the factor, for a response surface methods (RSM) design. How can I do that? Can I use the optimization screen in Design-Expert?"
Be very wary when extrapolating, but, by clicking the Sheet mode on the Factors Tool for the Point Prediction node in Design-Expert, you can enter levels that go out of the original bounds and generate predictions. Heed the advice that crops up in the software: *** WARNING - One or more factor value(s) is outside of the design space.
(Learn more about developing good RSM designs by attending the "Response Surface Methods for Process Optimization" workshop. For a description, see http://www.statease.com/clas_rsm.html. 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.)
5 - Simulation alert: Brew better beer by doing DOE
Bill Hathaway of MoreSteam.com offers
a new simulation* that allows you to try your hand at brewing better beer by
doing DOE. Click http://www.moresteam.com/betterbeer.cfm (simulator no longer available)
to do the fermentation for better or worse beer. I hope you do better than me
the first time through. I think I killed the taster. He got all choked up and
his face turned green!
*(The is the third DOE sim developed by MoreSteam.com. In my January e-mail I provided a temporary link to a beta version of a simulator that shoots virtual boulders from a trebuchet. Next came a pizza-making simulation to which I gave a link in my April DOE FAQ Alert. Stat-Ease has no financial interest in MoreSteam.com, but they use "DOE Simplified: Practical Tools for Effective Experimentation" (co-authored by me and my partner Pat Whitcomb) as their DOE textbook, which comes with a free trial copy of Design-Ease® software. For details on this book, see
6 - Workshop alert: "Experiment
Design Made Easy" will be in Philadelphia (plus a link to other locations
and/or other DOE workshops)
Our "Experiment Design Made Easy" (EDME) workshop will be presented in Philadelphia next month on October 8-10. However, if you need this basic training on DOE immediately, there's still room in our class here in Minneapolis coming up next week (September 10-12). See http://www.statease.com/clas_pub.html for schedule and site information on all Stat-Ease workshops open to the public, not just the EDME, but many others more or less advanced. 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.
7 - Events alert: A heads-up on DOE talks in Akron (9/11!), Chicago, and the Philadelphia area (Valley Forge)
I will be giving a talk at the International
Tire Exposition and Conference (ITEC) on Wednesday morning, September 11 in
Akron, Ohio. See http://www.itectireshow.com/2002/index.asp
(no longer available) for details. Here's the press release: "Come to "Statistical Design
of Experiments (DOE) for Tire Chemists and Engineers" and "meet the
expert" Mark Anderson, co-author of "DOE Simplified, Practical Tools
for Effective Experimentation." Mark is a principal of Stat-Ease, Inc.,
a world-renowned provider of consulting and training on DOE. He will distill
100 hours of their workshop material into the essential aspects needed to achieve
breakthroughs in process efficiency and product quality. The presentation will
be kept simple and fun with a minimum of mathematics in the spirit of the Stat-Ease
motto: 'Statistics Made Easy.'"
On Monday, September 30 I will be in Chicago at Vinyltec, sponsored by the Vinyl Division and the Polymer Modifier and Additives Division of the Society of Plastics Engineers (SPE). My talk is entitled "Achieving Six Sigma Objectives for Variability Reduction in Polymer Processing." For more information, see
http://www.4spe.org/seminarsconferences/conferences/vinyltec02_09.htm (Archives only go back to 2004).
On Friday, October 18 Pat Whitcomb
and Gary Oehlert will unveil a new class of designs that offer good value for
industrial experimenters who cannot afford to do more runs than necessary for
adequate response modeling. Their talk, entitled "Central Composite Designs
from Small, Optimal, Equireplicated, Resolution V Fractions of 2^k Designs,"
will be presented in the afternoon track of "Topics in Designed Experiments"
at the Fall Technical Conference (FTC) in Valley Forge, Pennsylvania. For details,
see http://www.cpid.net/46FTC_Information.htm (site no longer available).
Click http://www.statease.com/events.html 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!
8 - Reader contribution for anniversary of September 11th tragedy: The story of a scientist from India who made a new life in the USA
From: North Carolina
"Over the past months I have
forwarded various articles and accounts dealing with patriotism and why people
love or should love this country. Yesterday I had a personal experience that
was tremendous, so today I can tell you firsthand about this, not quote some
past or current writer or historical information. I hope you find this as inspiring
as I did.
It was on my final flight home from a business trip, leaving Philadelphia to go to Raleigh. I got a seat up in 1st class, and the guy next to me was a young man, quite dark skinned, appearing to be Indian in origin. However, when we struck up a conversation his English was perfect, unaccented, so I thought he must be second generation, born and raised here. We talked about business, the state of the economy, etc, but gradually moved to American culture and history.
He told me that as a native Indian, born in Madras and raised there, when he came to the USA for college, he'd read the Constitution and found it more moving than the religious texts he was studying. And after 4 years of college and finding life here wonderful, he did everything possible to find a way to stay. "After 4 years in this country they'd have had to shoot me and put a toe tag on me before they could get me on a plane back to India" he said.
He married a nice North Carolina girl, got a job in the pharmaceutical industry, and they proceeded to build a life, and have three children. And he applied for citizenship, and took 5 years to get through the process. The day he was going down to the town hall for the swearing in, he got up all excited in the morning, and his three kids, then ages 9, 7, and 5, noticed it and started asking questions.
"What's the big deal, Daddy?"
"Well, after a long time of waiting and working and hoping, today I am going to get something you kids were born with, so you take it for granted. You have always been Americans, although maybe you don't yet really know what that means. But today I will become an American, which changes all my life, and lets me move away from what I was before and become new and better in many ways. And it lets me take in life here more deeply than before, which is a wonderful thing."
So the kids thought for a while during
breakfast and then the oldest asked their mother "Mom, when Dad becomes
an American, will he then become more pink, like us?"
And the mother said "No, kids, Dad will just be a brown American, his skin won't change."
And the middle child then said "Well, will Dad start to eat fried chicken with us?" (He is still a rigorous vegetarian.)
And the mother said "No, Dad won't change how he eats, he will just be a vegetarian American."
Finally, the youngest child said "How about church? Won't Dad start coming to Mass with us now?"
And the mother said "No, Dad will still have his religious beliefs and heritage. He will just be a Hindu American."
At that moment the man thought to
himself, This is what is great about this country - you can become an American
and still be
And he said to me, then on the plane, "Every day I look at my family, where and how we live, all that we have, all the good my children have to look forward to in their lives, and I am just amazed at it all, and so thankful to God. I pray to God that whatever I am required to do to stay here and keep this life, He gives me the strength to do it."
I could only shake his hand, and pretend to look away to find the flight attendant, since I did not want him to notice my eyes filling up or hear my voice shake.
This is what we are, what we have, what so many others fought and died for, and it takes a man born in a small dusty town in Madras, with no running water and little sanitation, to spell it out. I wish to God more Americans understood what he understands. Pass this on as you will."
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. (http://www.statease.com)
Minneapolis, Minnesota USA
PS. Quote for the month - David Letterman
on USA Today stats
"USA Today has come out with a new survey - apparently three out of every four people make up 75% of the population."
(From "Inference," published by the American Statistical Association.)
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 http://www.statease.com/consult.html for resumes)
- Statistical advisor to Stat-Ease: Dr. Gary Oehlert (http://www.statease.com/garyoehl.html)
- Stat-Ease programmers, especially Tryg Helseth (http://www.statease.com/pgmstaff.html)
- 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 (see above)