Dear Statalisters,
I have conducted a full-interaction model where all the variables are interacted by the year dummy.
Afterward, I am estimating the predictions of my dependent variable at years 2006 and 2012. I applied the contrast command. However, I need to get the f-test for the differences over the years to understand if the change of the predictions over time are significant or not.
Kindly advise me.
Many thanks,
Maye
I have conducted a full-interaction model where all the variables are interacted by the year dummy.
Afterward, I am estimating the predictions of my dependent variable at years 2006 and 2012. I applied the contrast command. However, I need to get the f-test for the differences over the years to understand if the change of the predictions over time are significant or not.
Code:
global indepinter4 "(i.agegrp5 i.educ4 i.marital3 i.urban i.public c.hhsize i.qwealth)##(year2012)" xttobit $dep $indepinter4 if female==1 [iw=panel_wt_06_12], ll(0) margins i.agegrp5 i.educ4 i.urban i.public i.marital3 i.qwealth, over(year2012) post contrast (r.agegrp5 r.educ4 r.urban r.public r.marital3 r.qwealth)@year2012, nowald effects
Many thanks,
Maye
Comment