Subscribe Bookmark RSS Feed

Q-Q plot/Prob Plots

outlier

Community Trekker

Joined:

Jun 23, 2011

Hi, I remember seeing a probability plot of multiple parameters which was overlayed and normalised. This means you can see quickly which parameter is likely driving another parameter by virtue of the distribution shape.

E.g. I have 5 columns (say A,B,C,D,E) of continous data and would like to overlay them all on one probability plot. As the data has completely different scales it needs to be normalised. If A is my response, and I see that A & C has similar shape (B,D,E different) I conclude C is likely driving A.

Is there a standard plot in JMP to do this, i.e. to automatically normalise it..?