I had a question: on what basis do we pick a stcox vs streg model for time to event data?, and if we choose an streg model, how do we decide which distribution to go for? would it be on the basis of the graph from sts graph,haz?
also, what if the hazard curve is a little wonky (ups and downs), would we consider adding time splines?
also, what if the hazard curve is a little wonky (ups and downs), would we consider adding time splines?
