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

PS. Quote for the month: Beware of “error”!

1: FAQ: How to display the t-values of effects on a Pareto Chart

Original question from a Biochemical Research Engineer:
“On the Pareto Chart of effects Design-Expert® v8 software displays the t-value on the y-axis.  However, I cannot find an option that lets me see what the exact t-values are for the columns and it is not accurate to just read off the axis.  How can I get this data?”

Answer from Stat-Ease Consultant Wayne Adams:
“The t-values of the bars are the square root of the F values on the ANOVA.  That is one way to get what you want.  Another way is to hold the control (ctrl) key down and right-click on the white area of the Pareto chart.”


Pareto Chart of effects with t-values displayed

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

2: FAQ: Using the half-normal plot to select effects

Answer from Stat-Ease Consultant Shari Kraber:
“The tutorials are a great way to learn how to use Design-Expert.  Screen Tips and the Help System also provide great guidance and many answers to statistical questions.

Various statisticians have tried to develop algorithmic methods to model factorials, but none consistently work in a variety of situations, with both small and large numbers of effects.  It turns out that the human eye is better at detecting patterns than any automated process, at least for now.  Theory says that insignificant effects will follow a normal distribution, thus on a probability plot the insignificant ones fall on a nice straight line from zero, while the ones that are statistically significant ‘pop off’ the line.  It takes a bit of practice, but once you have worked through a few examples, you will pick off the significant effects very quickly.

For more details, see my write-up on “Over-Selection of Effects on the Half-Normal Plot” on page 3 of the September 2009 Stat-Teaser newsletter.”


Half-normal plot of effects with values displayed

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

• Experiment Design Made Easy (EDME)
• December 10-11*
  
  
• Response Surface Methods for Process Optimization (RSM)
  • December 12-13*

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

• Mixture Design for Optimal Formulations (MIX)
  • November 5-6**
    
    
• Advanced Formulations: Combining Mixture & Process Variables (MIX2)
  • November 7-8**

**Take both MIX and MIX2 in the same week 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 Shari at 1-612-746-2035.  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].

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

Please do not send me requests to subscribe or unsubscribe—follow the instructions at the end of this message. 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—Beware of error:

"There is often a confusion between the ordinary meaning of words and the specific mathematical meaning when the same words appear in a statistical analysis.  When I first began to work in the drug industry, one of my analyses included a traditional table of results, where one line referred to the uncertainty produced by small random fluctuations in the data.  This line is called, in the traditional table, the “error.”  One of the senior executives refused to send such a report to the U.S. Food and Drug Administration. “How can we admit to having error in our data?” he asked, referring to the extensive efforts that had been made to be sure the clinical data were correct.  He insisted I find some other way to describe it.  He would not send a report admitting error to the FDA.  I contacted H.F. Smith at the University of Connecticut and explained my problem.  He suggested that I call the line the “residual,” pointing out that in some papers this is referred to as the residual error.  I mentioned this to other statisticians working in the industry, and they began to use it.  This eventually became the standard terminology in most of the medical literature.  It seems that no one, in the United States at least, will admit to having error.”

—Anecdote by David S. Salsburg, p. 239, The Lady Tasting Tea, How Statistics Revolutionized Science in the Twentieth Century

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 sales and marketing director, Karen Dulski, and all the remaining staff that provide such supreme support!

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

DOE FAQ Alert ©2013 Stat-Ease, Inc.
Circulation: Over 6300 worldwide
All rights reserved.