Stat-Ease
> If you are having trouble viewing this email view it online.
 
Vol: 11 | No: 4 | Jul/Aug'11
Stat-Ease
The DOE FAQ Alert
     
 

Heads-up (below!)
The meaning of an interaction…

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, click here.

TIP: Get immediate answers to questions about DOE via the Search feature on the main menu of the Stat-Ease® web site. This not only pores over previous alerts, but also the wealth of technical publications posted throughout the site.

Feel free to forward this newsletter to your colleagues. They can subscribe by going to this registration page.

Also, Stat-Ease offers an interactive website—The Support Forum for Experiment Design. Anyone (after gaining approval for registration) can post questions and answers to the forum, which is open for all to see (with moderation). Furthermore the forum provides program help for Design-Ease® and Design-Expert® software. Check it out and search for answers. If you come up empty, do not be shy: Ask your question! Also, this being a forum, we encourage you to weigh in with answers! The following Support Forum topic provides a sample of threads that developed since my last Alert:

  • "How to best set up DOE with exact constraints"

To open yet another avenue of communications with fellow DOE aficionados, sign up for The Stat-Ease Professional Network on Linked In and start or participate in discussions with other software users.

 
Stats Made Easy Blog

StatsMadeEasy offers wry comments weekly from an engineer with a bent for experimentation and statistics. Simply enter your e-mail in the forwarding field at www.StatsMadeEasy.net  and get new StatsMadeEasy entries delivered directly to your inbox. Or, click this link to:

Subscribe with Feedburner

“Your StatsMadeEasy Blog brightens up a dreary work day...”
—Applied Statistician, Florida Smiley Face

Topics discussed since the last issue of the DOE FAQ Alert (latest one first):

Also see the new comments on my 4/2/11 blog-alert on “Supreme Court overturns tyranny of statistical significance” and other recent posts. Please do not be shy about adding your take about any news or views you see in StatsMadeEasy.  Thanks for paying attention.

 
     
 

If this newsletter prompts you to ask your own questions about DOE, please address them via e-mail to: [email protected].


-


Topics in the body text of this DOE FAQ Alert are headlined below (the expert ones, if any, delve into statistical details):

1:  Software Alert: Version 8.0.6 of Design-Expert software released (free update for licensed users of v8)
2:  Newsletter Alert: August issue of the Stat-Teaser explains how you can “Fly Your Way to Better DOE” via in-class experimentation
3:  FAQ: How can you explain the meaning of an interaction?
4:  Expert FAQ: Selecting effects via the half-normal versus a backwards regression: How do you explain discrepancies between these two approaches?
5:  Webinar alert: Basics of Response Surface Methodology (RSM) for Process Optimization, Part 1
6:  Events alert: ‘Heads-up’ on “Design Space via DOE Using Intervals to Manage Uncertainty” for industrial statisticians in Europe
7:  Workshop alert: See when and where to learn about DOE in the USA, Belgium and India
 
 
PS. Quote for the month: Insightful observations on intelligence
(Page down to the end of this e-zine to enjoy the actual quote.)


- Back to top -
 
 

1: Software Alert: Version 8.0.6 of Design-Expert software released (free update for licensed users of v8)

Newly-released version 8.0.6 of Design-Expert software is posted at this download site for free trial evaluation.  This web site also provides free patches to update older licensed versions of 8.0.

The release is primarily for maintenance, but it does provides new features, for example, improved color differentiation for design points displayed on model graphs.

View the ReadMe file for other features, installation tips, known ‘bugs,’ change history, and FAQs.

PS. Heads-up: 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 the “Check for updates on program start” option.


- Back to top -


2: Newsletter alert: August issue of the Stat-Teaser explains how you can “Fly Your Way to Better DOE” via in-class experimentation

Many of you have received, or soon will, a printed copy of the latest Stat-Teaser, but others, by choice or because you reside outside of North America, will get your only view of the August issue at this link.

It features an article by Consultant Brooks Henderson that explains how you can “Fly Your Way to Better DOE” via in-class experimentation.  He reports on results from the latest paper-helicopter fly-offs at South Dakota School of Mines & Technology—his alma mater.

This Stat-Teaser also provides an explanation on the “Power of DOE Using Qualitative vs. Quantitative Variables” contributed by Arved Harding—a “friendly neighborhood statistician” who hails from Kingsport, Tennessee.

Thank you for reading our newsletter.  If you get the hard copy, but find it just as convenient to read what we post to the Internet, consider contacting us to be taken off our mailing list, thus conserving resources.  (Note: You will be notified via the DOE FAQ Alert on new newsletter posts.)  In any case, we appreciate you passing along hard copies and/or the link to the posting of the Stat-Teaser to your colleagues.


- Back to top -


3: FAQ: How can you explain the meaning of an interaction?

Original Question:

From a Technology Improvement Head:
“Thanks so much for the webinar on How to Get Started with DOE, it was very clear and useful for me.  I have a question regarding interactions: How can you explain the meaning of the interaction?  I mean, is it correct to say that when you increase the value of your variable A (stirring) and decrease the value of B (pre-heat) you will have an additional effect on taste due to the interaction?  Is this a positive interaction?”

Answer:

From Stat-Ease Consultant Brooks Henderson:
“You are welcome.  I am glad you enjoyed the webinar.

Let me answer your question by going back to the example that I presented.  What happens with an interaction is that the graph for one factor will change slopes depending on the level of another factor.  Here are some pictures to show you what happens.  For the popcorn DOE let’s look at the taste response.  If we look at just a one factor plot of stirring (see first graph below), you don’t get the entire picture (note the warning in red).  On the first graph, I have factor B set to the low level.  In this case, the slope of the line is steeper.  In the second graph below, I have dragged factor B to the high level (by clicking and dragging the red bar on the factors tool below—see the green circle).  Here, with factor B at the high level (second graph), the slope of the line is not as steep.  That is due to the significant AB interaction.  The plot of A changes depending on the level of B.

One Factor Plot 3-1One Factor Plot 3-2
The effect of one factor (A) depends on the other (B)

In the two-factor interaction (2FI) plot I showed during the webinar, the two different slopes are shown on the same graph (see below).  This is the plot of the AB interaction.  Here, the black line indicates what happens to my response as we go from the low level to the high level of factor A while factor B is held at the low level (B- = 0.00 seconds).  The red line is what the factor A graph looks like when B is held at the high level (B+ = 360.00 seconds).  Again, the interaction just indicates that the graph for factor A will change depending on what level B is set to. The same is true for the factor B graph.  It will depend on what level of A you are looking at.

Interaction Graph
Interaction plot

So, to summarize, just be sure to look at the interaction graph when there is a significant interaction involving a factor.  Do not look at the graph for just that factor, because it does not show the whole story.  You need to look at the interaction graph.”

(Learn more about interactions by attending the two-day computer-intensive workshop “Experiment Design Made Easy.” Click on the title for a complete description.  Link from this page to the course outline and schedule.  Then, if you like, enroll online.)


- Back to top -


4: Expert FAQ: Selecting effects via the half-normal versus a backwards regression: How do you explain discrepancies between these two approaches?

Original Question:

From an Industrial Statistician:
For several designs that I have analyzed the half-normal and Pareto plots indicate that nothing is significant, but if you do a backwards regression on the effects list things show up as being significant.  I would like to communicate to my colleagues the correct way to interpret results, especially when two methods indicate that nothing is significant and the third method indicates that several terms are significant.  For example, try analyzing these responses from a 23 design (listed in standard order): 15.6, 18.2, 20.0, 40.4, 55.2, 55.4, 23.0, 41.2—the half-normal shows nothing significant but ANOVA lists all effects significant except for the ABC.”

Half-Normal Plot

Nothing looks significant on the half-normal plot

Anova
Everything is significant on the ANOVA

Answer:

From Stat-Ease Consultant Wayne Adams:
When effects are selected, they are moved from the error pool into the model pool.  The ANOVA compares the size of the model pool to the size of the error pool.  If the model pool is enough larger, the test comes up significant.  The key is you must only move explanatory terms from the error pool to the model.  If you move the largest error term, then the error pool is being underestimated and the tests can show false significance.  Backward regression algorithms start with everything (all terms) in the model before removing the least-significant term.  In your example this term is ABC.  The effect attributed to ABC is orders of magnitude smaller than everything else!  Given that this is the only thing in error, the test shows the model pool is significantly larger than the error pool, when the truth is plain to see on the half-normal plot (shown below): All of the effects should be part of the error pool, that is, nothing is significant.

Selection algorithms should only be used when you know something truly is significant in the model.  Even then, the selection needs to be backed up with a careful review of the ANOVA report that addresses these questions:

  • Are all the terms significant or required to support hierarchy?
  • Is the lack-of-fit insignificant?
  • Do the adjusted and predicted R-squared values come out positive and within 0.2 of each other?
  • Are the diagnostic plots OK?
  • Most importantly—does the model make sense to you?

The half-normal plot is the best method we’ve found for selecting the correct effects for the model.”

PS.  Check out an illustrated detailing of “Over-Selection of Effects on the Half-Normal Plot” by Stat-Ease Consultant Shari Kraber on page 3 of the September 2009 issue of the Stat-Teaser posted here.


- Back to top -


5: Webinar alert: Basics of Response Surface Methodology (RSM) for Process Optimization, Part 1

Response Surface Methods (RSM) can lead you to the peak of process performance.  In this intermediate-level webinar presented on Thursday, September 8th, 2011 at 2 pm CDT*, Stat-Ease Consultant Shari Kraber will introduce the fundamental concepts of response surface methods (RSM).

If you are new to RSM, this webinar is for you!  Stat-Ease webinars vary somewhat in length depending on the presenter and the particular session—mainly due to breaks for questions: Plan for 45 minutes to 1.5 hours, with 1 hour being the target median.  When developing these one-hour educational sessions, our presenters often draw valuable material from Stat-Ease DOE workshops.  Attendance may be limited, so sign up soon by contacting our Communications Specialist, Karen Dulski, via [email protected].  If you can be accommodated, she will provide immediate confirmation and, in timely fashion, the link with instructions from our web-conferencing vendor GotoWebinar.

*(To determine the time in your zone of the world, try using this link.  We are based in Minneapolis, which appears on the city list that you must manipulate to calculate the time correctly.  Evidently, correlating the clock on international communications is even more complicated than statistics!  Good luck!).


- Back to top -


6: Events alert: ‘Heads-up’ on “Design Space via DOE Using Intervals to Manage Uncertainty” for industrial statisticians in Europe

Stat-Ease Consultant Pat Whitcomb will detail “Design Space via DOE Using Intervals to Manage Uncertainty” at the European Network for Business and Industrial Statistics (ENBIS) annual gathering in Coimbra, Portugal on September 4-8.  Get details on the conference here.

On September 13 at IEEE’s Holm Conference on Electrical Contacts in Minneapolis I will present the Morton Antler Lecture on “Modern Design of Experiments (DOE)—75 Years of Advancements in Multifactor Test Methods”.  For more information click here.

Click here for a list of upcoming appearances by Stat-Ease professionals.  We hope to see you sometime in the near future!

PS.  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.


- Back to top -


7:Workshop alert: See when and where to learn about DOE in the USA, Belgium and India

Seats are filling fast for the following DOE classes.  If possible, enroll at least 4 weeks prior to the date so your place can be assured.  However, do not hesitate to ask whether seats remain on classes that are fast approaching!  Also, take advantage of a $395 discount when you take two complementary workshops that are offered on consecutive days.

All classes listed below will be held at the Stat-Ease training center in Minneapolis unless otherwise noted.

* Take both EDME and RSM in June to earn $395 off the combined tuition!

** Attend both SDOE and DELS to save $295 in overall cost.

*** Take both MIX and MIX2 to earn $395 off the combined tuition!

See this web page 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 [email protected].


- Back to top -

 

Please do not send me requests to subscribe or unsubscribe—follow the instructions at the very end of this message.
I hope you learned something from this issue. Address your general questions and comments to me at: [email protected].

Sincerely,

Mark

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


PS. Quote for the month—insightful observations on intelligence:


"
“Intelligence is not something possessed once for all.  It is in constant process of forming, and its retention requires constant alertness in observing consequences, and open-minded will to learn and courage in readjustment.”

—John Dewey, American empiricist, 1859–1952


Trademarks: Stat-Ease, Design-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 Stat-Ease software
Stat-Ease consultants Pat Whitcomb, Shari Kraber, Wayne Adams and Brooks Henderson
—Statistical advisor to Stat-Ease: Dr. Gary Oehlert
Stat-Ease programmers led by Neal Vaughn
—Heidi Hansel Wolfe, Stat-Ease marketing director, Karen Dulski, and all the remaining staff that provide such supreme support!

Twitter-SmileyFor breaking news from Stat-Ease go to this Twitter site.

DOE FAQ Alert ©2011 Stat-Ease, Inc.
Circulation: Over 5500 worldwide
All rights reserved.

 
  Subscribe