Hi,
I am running an interrupted time series analysis to assess the impact of a policy on medicine consumption. I have data for 5 countries & ~100 medicines for 24 time periods (~500 panels). All medicines/countries are impacted by the policy at the same time.
To use Ariel Linden's itsa command (ssc install itsa), am I correct in my understanding that the multiple groups options cannot be used in this situation? Ie, it can only be used to separate treated and non-treated panels?
If I were to use itsa, would I first have to aggregate the data over all countries/medicines for each time point?
If this is correct, it seems that I might lose valuable data/observations & I should perhaps use xtreg (or a similar panel regression method), with fixed effects for the countries and medicines, and include a policy variable (=1 from the year of policy implementation).
Thank you for any assistance.
I am running an interrupted time series analysis to assess the impact of a policy on medicine consumption. I have data for 5 countries & ~100 medicines for 24 time periods (~500 panels). All medicines/countries are impacted by the policy at the same time.
To use Ariel Linden's itsa command (ssc install itsa), am I correct in my understanding that the multiple groups options cannot be used in this situation? Ie, it can only be used to separate treated and non-treated panels?
If I were to use itsa, would I first have to aggregate the data over all countries/medicines for each time point?
If this is correct, it seems that I might lose valuable data/observations & I should perhaps use xtreg (or a similar panel regression method), with fixed effects for the countries and medicines, and include a policy variable (=1 from the year of policy implementation).
Thank you for any assistance.
Comment