BookmarkSubscribeRSS Feed
Choose Language Hide Translation Bar
ChrisLooi

Community Trekker

Joined:

May 29, 2017

Power Analysis in Design Evaluation (DOE)

Hi,

 

Under the Design Evaluation (after the DOE is created), there is a tab called Power Analysis per below photo.

 

Picture1.png

I understood the Power represents the probability of detecting a significant effect (if there is). For example here, this design have a probability of 57.2% to detect effect of X1 if it is a significant effect.

 

My questions:

1.  Understood the Coefficient is the "slope" or effect strength on the response. This is represented under column Anticipated Coefficient. Why does this value is being put as default "1" ?

 

2. For example, if I were to change X1 Anticipated Coefficient from "1 to 2", the Power will jump from 57.2% to 98.4%. Am I correct to say, if X1 slope is "2" then this design have the probability to detect this effect at 98.7%?

 

3. Does the Power Analysis takes into account the Scaling Factor? Having a slope of 0.0002 doesn't mean there aren't any strong correlation between Y and X.

For example : The X-axis data is recorded in nanometer instead of meter.

 

Thanks for advise.

 

B.r,

Chris Looi

1 REPLY 1
Byron_JMP

Staff

Joined:

Apr 26, 2012

Re: Power Analysis in Design Evaluation (DOE)

Go to this link:

https://www.jmp.com/support/help/14/power-analysis.shtml

 

scroll down to the Power Analysis for Coffee Experiment

It explains the interpetation of the values.

JMP Systems Engineer, Pharm and BioPharm Sciences