I have a question regarding my prior Latent Class simulation settings.
In HB I used the following settings for my simulation:
1. Study (n=800): 20.000 Itertions, 50.000 draws, 10 prior degrees of dreedom, 0.5 prior variance
2. Study (n=500):20.000 Itertions, 50.000 draws, 5 prior degrees of dreedom, 0.5 prior variance
Both studies are the same in attributes and levels (only different languages were used)
What would you recommend for LC Settings regarding: max. number of iterations, number of replications and min/max number of groups?
For the 1. Study I was thinking of using 4 groups and for the 2. Study max number of groups = 2.
What about the other factors such as iterations and replications? Does my HB settings have any impact on that? Can I take the default settings of Sawtooth?
Thank you very much!