Subscribe Bookmark RSS Feed

In Custom Design DOE, how do I disallow factor levels related to season?

paul_h_anderson

Community Trekker

Joined:

Jan 21, 2015

Say I’m going to test across both cold and warm weather. There are certain factors and levels that I cannot test during a particular season. Other factors and levels that I have no interest in testing during a particular season (but it would not damage anything). Assume I can estimate cold season as being the first half of my test, how do I control the design? I don’t think I can set season as a blocking factor, as JMP doesn’t let me disallow factor-levels where design . has blocking. Thanks!

1 ACCEPTED SOLUTION

Accepted Solutions
louv

Staff

Joined:

Jun 23, 2011

Solution

Paul,

Perhaps you have two designs to perform. One process where the design examines cold weather and one design that investigates warm weather. I only suggest this since I am not sure you can truly randomize the experiment and even if you  make season a "hard to change" factor using a split-plot could take some time to cycle through the cold and warm seasons?

The disallowed combinations can be handled quite elegantly coming in JMP 12:)

Lou V

2 REPLIES
louv

Staff

Joined:

Jun 23, 2011

Solution

Paul,

Perhaps you have two designs to perform. One process where the design examines cold weather and one design that investigates warm weather. I only suggest this since I am not sure you can truly randomize the experiment and even if you  make season a "hard to change" factor using a split-plot could take some time to cycle through the cold and warm seasons?

The disallowed combinations can be handled quite elegantly coming in JMP 12:)

Lou V

paul_h_anderson

Community Trekker

Joined:

Jan 21, 2015

Thanks for response LouV.

I tried setting up 1st half of test with all factors, no season as blocking factor. Then I augmented the design for 2nd half with adding disallow factors, making sure not to check blocking. It SEEMS to have worked.

Cheers!