How does JMP treat replicates if it is not specified in model?

Question

Hi all,

I did an experiment with 2 by 3 factorial design of treatments, so there are 6 treatment combinations total. Each treatment combination has 4 replicates. In model, I only specified the treatment effects and their interactions, and did not specify replicates as random effect. JMP still took average of the replicates. I wonder how JMP treated the replicate since it was not specified. Do I have to rerun the analysis with replicate as random effect?

Thank you very much!

Amy

txnelson · Answer

I assume that JMP is adding the replicates to the calculations for the appropriate other effects in the model, and the replicate effects will be added to the error(non model) effect.