From: annamb5 <anabras@...>
Subject: [DOE_DG] Factor interaction in L9
To: DOE_DG@yahoogroups.com
Date: Wednesday, March 11, 2009, 4:22 PM

Hello,

I have some new questions about factor interactions that I would like to expose to you.
In my problem I have identified 4 factors (A,B,C and D) with three levels each. So I adopted a L9 array.
The first level of factor D depends on factor A and C.

According to DOE using Taguchi approach from Ranjit Roy, the trial conditions are unaffected by interactions and I understand that it means I may start the experimental research (without previously knowing if there are or not interaction factors).

I believe that if there are interactions between the factors that I want to study, its effect is really low.

However I do not know if am I doing it right for factor D since its first level depends on A and C…

Can I proceed with the 9 tests without knowing those answers? These tests are expensive and I also do not have enough time to spend.

Is it possible to use QT4 to identify interaction factors and its significant (or not) after all tests were made using L9 array?

Another question:
both the process of designing the experiment and the statistical analyses of the results are considered in the Taguchi approach, which is based on the use of orthogonal array, associated with the analysis of variance (ANOVA), and the significance test with F statistic.
Why is it used the F statistic and not other type of significance test?

I would appreciate if you could answer to my questions.

Thank you very much.

Ana B.