DOE FAQ Alert


DOE FAQ Alert Electronic Newsletter

Issue: Volume 2, Number 2
Date:
February 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, go to http://www.statease.com/doealert.html.
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 as an appetizer: http://pbskids.org/dragonflytv/, which offers ideas and experiments for children aged 9-12. The web site is done in conjunction with a new USA public TV series called "Dragonfly," showing "real kids doing real science," such as trying to get electricity from a cow.(?) Click on the link labeled "investigate this" and try the first experiment on the list: Measure the Earth. Based on measurements of shadows, a Greek thinker named Eratosthenes did this more than 2000 years ago. Now, with the aid of Dragonfly, you too can measure the earth! It would be interesting to know the accuracy and precision of this measurement method.

Here's what I cover in the body text of this DOE FAQ Alert:

1. Alert: Free upgrade available for V6.06 of Design-Expert® or Design-Ease® software (if not currently a user, get the free 30-day trial version)
2. FAQ: Explaining R-squared, the predicted form versus other variations on this statistic
3. Workshop alert: Do you know the way to San Jose?
4. Book alert: Giving away more statistical texts (free); New editions out for "Response Surface Methodology" and "Experiments with Mixtures", plus other new books available via the Stat-Ease web site

PS. Quote for the month - H. G. Wells on statistical thinking


1 - Alert: Free upgrade available for V6.06 of Design-Expert or Design-Ease software (if not currently a user, get the free 30-day trial version)

If you (as an individual user) own a permanently licensed copy of version 6 of Design-Ease (DE6) or Design-Expert (DX6), go to http://www.statease.com/soft_ftp.html#dx6updt for downloads that will patch your software with the latest enhancements. We recommend you do this even though the changes may affect only a few users. To see what got added or fixed, click the "Read Me" link for either DE or DX (whichever program you will be updating).

If you own networked software that needs updating, you must call Stat-Ease customer service at 1.612.378.9449. We do not post patches for networked software on the Web. Be prepared to provide your serial number. We will then send you a replacement CD-ROM to re-install on your network. As noted above, you should see the "Read Me" file for details on the latest changes, but of particular interest to network administrators is this fix: "Network installations that use Universal Naming Convention (UNC) path names may produce an error when opening Help. This is a problem inherent in the Win2000 help format that displays the index and content side-by-side in the same window. The cure for that was to map a drive letter to the program location. For administrators who want to stick with pure UNC names, we have added a standard single-pane version of the help."

Before updating or buying version 6 of Design-Expert, feel free to download a fully-functional, but 30-day limited, version of the software from http://www.statease.com/dx6trial.html. Users of Design-Ease or earlier versions of Design-Expert (V5 or before) should consider upgrading their software to DX6. See why you should do it at http://www.statease.com/dx6descr.html. Then call Stat-Ease for a quote. After validating your current serial number, they will give you a special upgrade price.

PS. I recently bought a new Dell laptop with Windows and Office XP. I've been using the new version of Design-Expert for some time now with no problems on this new Microsoft operating system and the latest edition of Word and Excel. Mark


2 - FAQ: Explaining R-squared, the predicted form versus other variations on this statistic

-----Original Question-----

From: New York

"I have a question about the predicted R-squared statistic. I just read someone's comment [another publisher of statistical software] that it measures goodness of prediction and should be used to choose a model in place of R-squared or adjusted R-squared. Does that sound like an OK way for me to explain that to others?"

Answer:

Yes, I think you hit the nail on the head. However, I advise that before getting into the advanced forms of R-squared you go back to the basics on this powerful statistic. A good place to start might be a newsletter article written by my partner Pat entitled "Don't Let R^2 Fool You" (see page 2 of September, 1997, "Stat-Teaser" at http://www.statease.com/news/news9709.pdf). A copy of this article is included in our Handbook for Experimenters" (page 3-1) along with lots of other useful information about DOE.* For you readers, I offer this link to a rather cynical cartoon on a related issue - correlation: http://www.strange-matter.com/li/li037.jpg (link no longer available).

After gaining an appreciation for what the raw R-squared can or cannot do, it's good to get into the adjusted and predicted forms of this statistic. We use these for scoring models in the response surface and mixture design portions of our Design-Expert software. This is described on page 6 in section 12 of our "User's Guide" (available on the Web at http://www.statease.info/dx6files/manual//DX12-Details-Analysis.pdf.)

Watch out when using predicted R-squared (or adjusted R-Squared): It may go negative! This occurs when the underlying predicted residual sums of squares (PRESS) gets inflated beyond the level of the corrected (for the mean) total SS. (For definition of PRESS, see the aforementioned reference to the DX "User's Guide"). We illustrate how predicted R-squared can become negative in our class "Response Surface Methods for Process Optimization" with the graphic shown via this link: http://www.statease.com/misc_data/negpress.pdf. Notice in the graphic how much variation occurs in the slope when you take each point out and re-fit the line. The accumulated PRESS is far more than the total SS (corrected for the mean), so R-squared predicted goes negative. For this particular set of data, a quadratic model is needed for a better fit, one that would produce a positive predicted R-squared value.

*Purchasers of Design-Ease and Design-Expert software get a free copy of the Handbook when they register their license. Copies can be purchased for $5 at our e-commerce book site http://www.statease.com/prodbook.html.

[Learn more about developing good predictive models by attending the "Response Surface Methods for Process Optimization" workshop (see http://www.statease.com/clas_rsm.html).]


3 - Workshop alert: Do you know the way to San Jose?

We've got a full house at our Experiment Design Made Easy (EDME) workshop in San Jose this week (Feb. 5-7), so we scheduled another session for May 7-9. This replaces the class originally scheduled for Detroit. We also present EDME's in Minneapolis (our headquarters) in April (9-11) and June (4-6).

If you need more advanced training on DOE, consider coming to Minneapolis for:
- Robust Design, DOE Tools for Reducing Variation (a must for Six Sigma!), March 12-14 (Prerequisite: knowledge of RSM)
- Response Surface Methods (RSM) for Process Optimization, April 16-18

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 at our website 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.


4 - Book alert: Giving away more statistical texts; New editions out for "Response Surface Methodology" and "Experiments with Mixtures" plus other new books available via the Stat-Ease site

We recently cleaned house at Stat-Ease and found a number of old, but still relevant, statistical texts that we no longer use in our workshops. This month I am giving away 6 paperback copies of Wheeler and Chamber's classic "Understanding Statistical Process Control," first edition (1986). The first 6 people* who e-mail their request to me will get a copy of this book. Please note the shipping address in your e-mail.

Last month I offered a similar number of Kume's "Statistical Methods for Quality Improvement" book. It took less than an hour for these to be spoken for!

*(I want to spread the wealth of information, so those of you who received a book from the last issue of DOE FAQ Alert won't be eligible this time.)

PS. I just got new editions of two books which I highly recommend you (or your company's technical library) buy:
- "Response Surface Methodology," 2nd Edition, by Myers and Montgomery, John Wiley & Sons, 2002
- "Experiments with Mixtures," 3rd Edition, by Cornell, John Wiley & Sons, 2002

Purchase these books at http://www.statease.com/prodbook.html. In addition to the new editions noted above, we recently added two
new books to our e-commerce site:
- "Engineering Statistics," 2nd Edition, by Montgomery, Runger and Hubele, John Wiley & Sons, 2001
- "Statistical Intervals: A Guide for Practitioners," by Hahn and Meeker, John Wiley & Sons, 1991

We recommend these texts for our new "Statistics for Technical Professionals" workshop (http://www.statease.com/clas_stp.html).


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

[email protected]

Mark J. Anderson, PE, CQE
Principal, Stat-Ease, Inc. (http://www.statease.com)
Minneapolis, Minnesota USA

PS. Quote for the month - H. G. Wells on statistical thinking:

"Statistical thinking will one day be as necessary for efficient citizenship as the ability to read and write."

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 (See above)

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

Statistics Made Easy

©2002 Stat-Ease, Inc. All rights reserved.