Multi-grouping GRR analysis JMP equations/statistics with bayesian estimate

[#1]

Hi,

I am trying to reproduce the GRR results in which there is more variance subgroups beyond just operator, and sample variance. e.g subgroup3,subgroup4.

With the default 2group operator/sample variance using anova crossed GR&R, i was able to reproduce the results using the anova method following the explanation. https://sixsigmastudyguide.com/repeatability-and-reproducibility-rr/

I noticed a message when i try variance analysis with more groups that the GRR analysis switch to using bayesian estimates because of negative REML variance. Was wondering if there is code or formula of how to calculate the variance using the bayesian variance component analysis?

[#2]

I don't know why it does that, but here's how to keep the bayesian estimate from coming up.

In preferences (file>preferences) pick this option

Byron_JMP_0-1685995227915.png