Hi!
I'm interested in changes in network size before and after widowhood.
I modeled three variables, f1 for linear change, f2 for quadratic change before widowhood; and f3 for quadratic change after widowhood.
I ran xtmixed and both f2 and f3 were statistically significant, indicating that the increase in network size tapered off with time after widowhood.
If I want to create graph that looks like this, showing the quadratic change before and after widowhood, what would be my next step?
Thank you!
I'm interested in changes in network size before and after widowhood.
I modeled three variables, f1 for linear change, f2 for quadratic change before widowhood; and f3 for quadratic change after widowhood.
Code:
gen f1=year-widowhood_year gen f2=(year-widowhood_year)^2 if year<= widowhood_year gen f3=(year-widowhood_year)^2 if year>widowhood_year
Code:
xtmixed size income retired health f1 f2 f3 || id : f1 age educ female
Thank you!