0

Considering the linear mixed model, $y=X\beta+Zb+e$.

After parameter estimation, we want to compute the MSE of the model. First, we compute the BLUP of b. Then $MSE = ||y-X\beta^*-Zb^*||^2$.

Is the formulation of MSE right? I'm not sure, because I didn't find any formula of MSE of linear mixed model. Is there anyone who can point me to the right reference or tell me what the MSE of linear mixed model is? Thank you so much!

Bing
  • 21
  • 1

1 Answers1

1

MSE is simply average squared residual error, so the formula is

$$ \frac{1}{N} \sum (y - \hat y)^2 $$

I do not think that there are any special corrections for LMM's.

Community
  • 1
Tim
  • 108,699
  • 20
  • 212
  • 390