The Elsmar Cove Wiki More Free Files The Elsmar Cove Forums Discussion Thread Index Post Attachments Listing Failure Modes Services and Solutions to Problems Elsmar cove Forums Main Page Elsmar Cove Home Page
Google
  Web Elsmar.com
*Please be aware that SOME RECENT forum threads may not yet be indexed by Google.

View Full Version : DOE - Design of Experiments - Shainin


Andrews
17th December 2002, 10:05 AM
CAN SOMEONE PLEASE DIRECT ME TO A WEBSITE THAT COVERS DESIGN OF EXPERIMENTS.
HOW MANY OF YOU USE THIS TO QUALIFY SPECIAL PROCESSES.
WHAT APPROACH DO YOU ADOPT TO QUALIFY SPECIAL PROCESSES

Bill Ryan
17th December 2002, 01:13 PM
Andrews,

There are any number of DOE websites. Therre are also modifications on the "DOE" formats, depending on which statistician you wish to "adopt". If you type "Design of Experiments" into your Search Engine, you'll be able to pick and choose which one suits you.

I'm relatively new here at the Cove, but you might try a Cove "Search" in case there has been discussion on the subject (I would doubt you would get the detail, but possibly a good overview of the power of DOE).

I'd like to give you a specific link but I don't know what level of information you are looking for. There are also numerous texts available - all with varying levels of topic "understanding" needed.

Sorry I can't be of more definitive help. My company took all our Technicians and Engineers through "In-house" training using George Box's methodolgy (and it has served us well!!), but we also use Taguchi and Shainin (depends on what the responsible engineer is used to).

I wish you luck!!!

Bill

Andrews
16th July 2003, 10:16 AM
What I was looking for was some website on Shainin DOE.I heard that Shainin DOE is faster in getting the results than Taguchi.Is it true? What is your opinion?

Ravi Khare
16th July 2003, 02:37 PM
Each of the DOE methods have their own merit.
Shainin DOE is the simplistic version.
Taguchi DOE comes useful if you are looking at Parameter Design and Tolerance Design. The concurrent statistic of S/N ratio is used in Taguchi methods. Taguchi Methods are effectively used in evaluation of Robustness and Robust Design.
Ronald Fisher's Classical DOE emphasizes ANOVA that can isolate real effects from the within treatment noise.
The best course of action would be to use the best combination of all the methods based on the demand of the problem.
As someone has rightly said on these forums; its Horses for Courses.

Groo3
21st August 2003, 11:52 AM
I don't have much experience with DOE, but we recently had to set up a Design of Experiments with one of our suppliers in order to satisfy a customer requirement. We were able to use Statgraphics to help set up a basic DOE on three variables. Regardless of whether you choose to go with this software package or one of the many other DOE software packages out there, you will probably need a day or two of training on how to use the software.

If you are just looking to research sites which cover DOE, there are quite a few sites on the internet... A Google search on "Design of Experiments" resulted in 1,770,000 pages found... I searched through several of the initial sites and came across the following informational / reference pages which may help?

http://www.mathworks.co.uk/access/helpdesk/help/toolbox/stats/doe.shtml

An article on DOE which references several DOE Software vendors:
http://www.aip.org/tip/INPHFA/vol-3/iss-3/p24.pdf

Unfortunately, when you conduct a Google search, it's hard to sort the list of DOE Software (or Excel Add-in) sites from the DOE Course offerings, seminars, basic DOE informational / reference sites, etc. An ASQ Search on the other hand, may provide you with the references you are looking for? Hope this helps you in your search?

Atul Khandekar
23rd August 2003, 09:17 AM
Good Links! Thanks.

ben sortin
10th October 2003, 11:40 AM
CAN SOMEONE PLEASE DIRECT ME TO A WEBSITE THAT COVERS DESIGN OF EXPERIMENTS.
HOW MANY OF YOU USE THIS TO QUALIFY SPECIAL PROCESSES.
WHAT APPROACH DO YOU ADOPT TO QUALIFY SPECIAL PROCESSES

Read "Statistics for Experimenters" by Box, Hunter, and Hunter. The publisher is John Wiley & Sons. Only the best web sites (designed to take your money) would quote from this stellar reference. Did I mention that "cover to cover" is the only way?

Confounded with all main effects clear,

Ben Sortin

Atul Khandekar
10th October 2003, 11:44 AM
Read "Statistics for Experimenters" by Box, Hunter, and Hunter. The publisher is John Wiley & Sons.
Excellent reference, Ben (wonder why I didn't think of it before! :) ) and welcome to the Cove..