Subscribe Bookmark RSS Feed

JMP10 DoE Lack of Fit

outlier

Community Trekker

Joined:

Jun 23, 2011

Can somebody help me interpret the meaning of the lack of fit test in the DOE platform?  In the help it mentions this is the delta of the residual model error and pure error.. so what is a "good" value?  Shouldn't both these values ideally be zero?  What is the test is significant (<0.05) does this mean the model is likely not a good fit?  I have significance of around 0.09.  The Rsquared (whole model) fit is 99.7%.