Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

3rd power parameter DOE

Oct 15, 2020 7:21 AM
(364 views)

Hi JMP ers,

I have 4 factors and try to make a model with main effects, interactions, 2nd order and 3rd order power.

I make design with recomended optimality criteria. When i checked color map correlations i can always see x1 and x1*x1*x1 correlation near to 1. How can make a DOE that does not have this problem?

18 REPLIES 18

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: 3rd power parameter DOE

**Customer Design** platform, in the **Model**, highlight the effects you want. If it says **If Possible**, you can change that to **Necessary**.

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: 3rd power parameter DOE

I tried but got same correlation again.

Did you check your correlation picture? Do i miss something diffrent?

Did you check your correlation picture? Do i miss something diffrent?

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: 3rd power parameter DOE

Ella,

What you are seeing is a result of the coding of values for the factors. -1^3 = -1, 1^3 = 1

So if you use "actual values" you will see the correlation matrix without those terms correlated.

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: 3rd power parameter DOE

i got same picture with real values

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: 3rd power parameter DOE

correlation between

x1, x1*x1*x1 - 0.92

x2, x2*x2*x2 - 0.93 etc.

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: 3rd power parameter DOE

Created:
Oct 16, 2020 4:25 AM
| Last Modified: Oct 16, 2020 6:07 AM
(220 views)
| Posted in reply to message from Ella 10-16-2020

The X^3 vs X plot has a correlation of 0.916 and the density ellipse is elongated. Find 4 points on that curve that are not correlated. I don't think you can do it. Those terms will always be correlated.

For a contrast, I have also included a plot of X^2 vs. X. The correlation is -0.019 and the density ellipse is almost circular. You can get correlations close to zero between these terms.

Dan Obermiller

Highlighted
##

I'm sorry, my understanding is that you will always get that correlation due to the coding that goes on "behind the scenes". What I meant to say, AFAIK, the x and x^3 will actually not be correlated if the appropriate designs selected and the actual values are used. The correlation matrix will always show the correlation due to the coding.

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: 3rd power parameter DOE

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: 3rd power parameter DOE

Created:
Oct 16, 2020 8:03 AM
| Last Modified: Oct 16, 2020 8:07 AM
(176 views)
| Posted in reply to message from statman 10-16-2020

Coding is actually used to REDUCE collinearity. But coding can only do so much. My graphs may not be the best way to illustrate, but consider this:

X1 X1^2 X1^3

100 10000 1000000

100 10000 1000000

200 40000 8000000

200 40000 8000000

150 22500 3375000

There is a correlation between X1 and X1^2 (0.997). But if you code to -1, +1, the correlation will drop to 0. This does not happen when cubing because, as you pointed out, -1^3 = +1 and +1^3 = 1, but the correlation on the original scale is large, too.

Dan Obermiller

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: 3rd power parameter DOE

I spoke to a JMP DOE developer yesterday (JMP summit) and he confirmed the plots would always show the correlation, but in actuality with the correct design, there is no correlation.