DFFITS plot for Standard Least Squares

[#1]

Hello everyone,

I'm searching for a possibility to easily display DFFITS plots for Standard Least Squares Models. I usually consider these plots (if I use different software) to investigate whether I have conspicuities. I usually have more than one response for the same DoE factor settings. Sometimes it happens that a run is an outlier in all DFFITS plots (but not suspicous at all in the plots JMP usually offers as Studentized Residuals, Normal Quantile etc.) which indicates that something went wrong here.

So this would really be a  great advantage if there exists some solution.

Best,

Anja

[#2 · Accepted solution]

I submitted this utility/demonstration years ago. It provides the information that you want. It works with only first-order terms, though.

[#3]

 Thank you very much! This helps really a lot. 

If anyone already modified this to also work with higher-order terms and/or variable selction included I would be glad to hear.