Hi everyone,
I'm rather new to STATA and ran into a puzzle I just can't seem to figure out.
For a panel event study (eventdd) I want to define an event which is the rapid, strong and sustainable increase of an indicator - to study this increase's event on my DV.
I want to study the effect of improvements in general, but there are too many improvements for each country in the observation period, so the effect of one improvement would lag into the next improvement. Therefore, I want to construct this "rapid improvement" binary which is defined as a 3p increase within 3 years which does not drop back down immediately in the next year or two (or only drops max. 1p after 2 years).
How would I go about constructing such an event dummy?
For background. I'm using the eventdd command, so I need to generate a variable that identifies the specific year (approx.) where the jump happens, for me to then generate the variables counting lags and leads.
I would also like to know how many of these jumps occur for each country, I'm hoping that it's only max. one per country but I will need to check that.
I very much hope this forum will be able to help me.
All the best,
Fabian
I'm rather new to STATA and ran into a puzzle I just can't seem to figure out.
For a panel event study (eventdd) I want to define an event which is the rapid, strong and sustainable increase of an indicator - to study this increase's event on my DV.
I want to study the effect of improvements in general, but there are too many improvements for each country in the observation period, so the effect of one improvement would lag into the next improvement. Therefore, I want to construct this "rapid improvement" binary which is defined as a 3p increase within 3 years which does not drop back down immediately in the next year or two (or only drops max. 1p after 2 years).
How would I go about constructing such an event dummy?
For background. I'm using the eventdd command, so I need to generate a variable that identifies the specific year (approx.) where the jump happens, for me to then generate the variables counting lags and leads.
I would also like to know how many of these jumps occur for each country, I'm hoping that it's only max. one per country but I will need to check that.
I very much hope this forum will be able to help me.
All the best,
Fabian
