Multi-state time-varying covariates
Posted: Tue Feb 19, 2019 9:21 amHi all,
Apologies if this is a stupid question, but from reading the MARK book I was wondering if there's a way to use time varying individual covariates in the multistate model without discretizing them as suggested in 11.6.1. I understand I could set the covariates up in a separate matrix (not using the multi-state approach), but then I would need to include these for the timepoints the animals were not seen in. I feel like for the type of covariates I'm trying to add to my ORDMS model, classes may be too subjective.
For a bit of context my individual time varying covariates are number of injuries which are either mild or severe.
Thanks so much in advance,
Jess
Apologies if this is a stupid question, but from reading the MARK book I was wondering if there's a way to use time varying individual covariates in the multistate model without discretizing them as suggested in 11.6.1. I understand I could set the covariates up in a separate matrix (not using the multi-state approach), but then I would need to include these for the timepoints the animals were not seen in. I feel like for the type of covariates I'm trying to add to my ORDMS model, classes may be too subjective.
For a bit of context my individual time varying covariates are number of injuries which are either mild or severe.
Thanks so much in advance,
Jess