Multi-state model averaging with many covariates

[thaljoha]

Hi,

I am running into issues when I try to model average transition probabilities (pregnant to not pregnant) in a multi-state model. I have a 25 year data set with two individual covariates (weight and age) and 4 environmental covariates (which I coded as individual covariates in the INP file). I have a covariate for each year of the data set so I end up having 25yrs x 6cov= 150 covariates. When I try to model average the psi in respect to an individual covariate, I receive an error message/ warning because I have more than 8 covariates. Any suggestions?

I ran models with no psi variation through time (using an intercept term only) and some where psi was time dependent.