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, 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 http://www.statease.com/doealertreg.html. If this newsletter prompts you to ask your own questions about DOE, please address them via mail to: StatHelp@StatEase.com.
Here's an appetizer to get this Alert off to a good start: For evidence (?) of life on Mars, check out this web site: http://www.space.com/scienceastronomy/mars_life_050216.html. Either this is a classic case of apophenia (see last month's DOE FAQ Alert for a definition of this word) or it's a plant by the PR people for the upcoming cinematic re-make of "War of Worlds" (see http://www.waroftheworlds.com/).
Here's what I cover in the body text of this DOE FAQ Alert (topics that delve into statistical detail are designated "Expert"):
1. FAQ: Percent contribution of a factorial effect
2. Free books (US/Canada only): Enter a drawing for "Engineering Statistics" and the newly-published "RSM Simplified" books
3. Reader contribution: DOE for designing a race car
4. Info Alert: Case study authors wanted
5. Events alert: Link to a schedule of Stat-Ease appearances
6. Workshop alert: Experiment Design Made Easy is coming to Philadelphia
PS. Quote for the month: See an astounding statistic on how few Americans truly understand how to properly conduct a scientific experiment.
To: Stat-Ease Consultant Shari Kraber
From: Software user in Alabama
"Shari, You were a tremendous help in explaining a question I had two months ago about the half-normal lots output by Design-Expert so I thought I'd try another question. When looking at the Effects List (View, Effects List) under the Analysis/Effects, what do the three columns of data ("Stdized Effects," "Sum of Squares," %Contribution) actually represent? For example, is it fair to say that %Contribution quantifies the amount to which a factor affects the model?"
Answer (from Shari):
"Without going into the math behind the effects list (I have to leave that for a regression class!), here are some conceptual definitions:
—Effect (in a two-level factorial design): This is the change in the response as you vary the factor from it's low level to it's high level. Obviously, larger effects are going to be more significant, but we have to use other tools, like the half-normal plot to determine the split between the significant and the non-significant effects.
—Sum of squares: Generally, you can think of sum of squares as the amount of information in the design that is attributed to that factor or term. Higher sums of squares correlate to higher effects.
—Percent Contribution: This is the percentage of the sum of squares for that term relative to the total sum of squares. Again, for a two-level factorial, this will correlate with the other columns. However, for general factorials (for example a 3x4x5) the column is biased and shouldn't be used. If a two-level factor contributes 50%, then that term accounts for half the total information in the data generated by the DOE.
If all responses do not get collected (missing data) and/or the factor levels are not set as originally planned (botched), these statistics may not be completely accurate. Also, at this stage of the analysis, effects have not yet been tested to see if they are statistically significant. For example, an effect that contributes a seemingly large amount, say 30%, may not be significant if the normal variation is large. Therefore, analysis of variance (ANOVA) must be employed to assess the significance of each effect. When calculating ANOVA, Stat-Ease software accounts for any anomalies in the design. The information on the effects screen should be only used as a guide for determining which terms to put in the model and which to leave in error, in order to calculate the ANOVA."
For a more detailed writeup on contribution, including examples, refer back to my June 2002 Alert (Volume 2, Number 6 posted at http://www.statease.com/news/faqalert2-6.html) where in the #2 FAQ I discussed "How to quantitatively compare effects from a two-level factorial design."
(Learn more about interpreting effects by attending the three-day computer-intensive workshop Experiment Design Made Easy." See http://www.statease.com/clas_edme.html for a course description. Link from this page to the course outline and schedule. Then, if you like, enroll online.)
Simply reply to this e-mail before April 1st if you'd like a chance at one of four free copies of "Engineering Statistics" by Montgomery, Hubele and Runger. Second Edition (Wiley, 2001). (Sorry, due to the high cost of shipping, this offer applies only to residents of the United States and Canada.) For workshops on this topic,* Stat-Ease now uses a different edition.
A drawing will also be held for a free copy of "RSM Simplified" autographed by the authors (myself and Stat-Ease consultant Patrick Whitcomb). For details on this newly-published book on response surface methods for process optimization, see http://www.statease.com/rsm_simplified.html, and/or copy and paste this path into your Web browser to see the publisher's PR: www.productivitypress.com/client/client_pages/pressprNov04-1.cfm.
For a complete list of books offered by Stat-Ease, go to its e-commerce site at http://www.statease.com/prodbook.html.
*(Learn more about engineering statistics by attending the three-day computer-intensive workshop Statistics for Technical Professionals." See http://www.statease.com/clas_stp.html for a complete description. Link from this page to the course outline and schedule. Then, if you like, enroll online.)
From: John J. Russell, Professor, Dept. of Mechanical Engineering, University of New Mexico
"I am very interested in obtaining Design-Ease or Design-Expert software for my Formula FSAE race car design class. I had a student illustrate the use of Design of Experiments using your trial product last year. We were easily able to home in on the design parameters most influential in the overall design. Here is our team web site: http://me.unm.edu/~fsae/teams/2005/. (Updated 3/07: Link no longer available.) We have been getting better each year finishing 39/140 in FSAE competition last year."
Professor Russell is now licensed to use Design-Expert for educational purposes. To see the presentation by his student, click http://www.statease.com/pubs/doeinracing.pdf. Also, see this related application for DOE contributed by Professor Mark Rusco of Ferris State University's Applied Technology Center in Michigan: http://www.statease.com/pubs/dragracing.pdf.
Gain technical recognition in your company and scientific field. Share a DOE with our writer at: Mailto:RABURNHAM@PublicationCoordination.com. Rename proprietary factors if needed to maintain confidentiality. A thirty-minute interview, some e-mails, a review, and it's done!
Stat-Ease plans on presenting its 3-day, computer-intensive Experiment Design Made Easy (EDME) workshop on March 29-31, 2005 in Philadelphia, PA. Enrollment will be closed out very soon, so do not delay if you'd like to get signed up for this EDME class: Call workshop coordinator Sherry Klick now at 612.378.9449 extension 18.
See http://www.statease.com/clas_pub.html for 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 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.
I hope you learned something from this issue. Address your general questions and comments to me at: Mark@StatEase.com.
Mark J. Anderson, PE, CQE
PS. Quote for the month—An astounding statistic on how many US adults understand the nature of a truly scientific experiment:
"A recent survey of US adults conducted by the National Science Foundation found that only 27 percent could acceptably describe the nature of scientific inquiry and explain what an experiment or a hypothesis is. In other words, most Americans don't grasp how science works or what separates it from pseudo-science. They sit on juries, yet they don't understand the nature of evidence."
—Richard Tresch Fienberg in a column titled "Evolution: We Can't Sit Idly By" in the April 2005 "Sky and Telescope" magazine, p. 8.
Trademarks: Design-Ease, Design-Expert and Stat-Ease are registered trademarks of Stat-Ease, Inc.
Acknowledgements to contributors:
in previous FAQ DOE Alert e-mail newsletters?
Click here to add your name to the FAQ
DOE Alert newsletter list server.
2021 E. Hennepin Avenue, Ste 480
Minneapolis, MN 55413-2726
p: 612.378.9449, f: 612.378.2152