Anonymous posted on Thursday, May 12, 2005 - 12:38 pm
I have a quick question. I've read your information concerning montecarlo simulations of SEM models. I'm wondering at about what level should I be concerned about parameter and standard error bias if I am trying to recover my SEM model using Monte Carlo simulations?
bmuthen posted on Thursday, May 12, 2005 - 6:40 pm
Hard to say in general, but something of the order of less than 5% bias in point estimates and less than 10% bias in SEs seems ok.