Hello everyone,
I would appreciate your advice on the matter of data visualization.
I am running a DD regression on the effect of legislation on millions of firm observations (panel, app. 10 mil. obs, 3 mil. firms, 10 years (missing values)).
The regression tells me that the effect of DD coefficient, an interaction of Treatment and Time, is positive and significant (p-value is 0,0000000) and the effect is app. 20 % increase.
The model looks like this:
The problem is that I would like to visualize the data, but when I do means/medians for individual years and Treatment (0/1), it seems insignificant or even negative.
I run some more regression and I saw that if I run the same regression without dummies accompanying DD coefficient (see code below), the DD coefficient is significant and negative.
This one is obviously biased as the dummies are not included, however it is what I can see when I try to visualize the data using means/medians for individual years and Treatment.
Can you please help me how to visualize the data in a case of this DD or DDD regression so that I can accompany my work also with visualization?
Thank you very much
Vojtech
I would appreciate your advice on the matter of data visualization.
I am running a DD regression on the effect of legislation on millions of firm observations (panel, app. 10 mil. obs, 3 mil. firms, 10 years (missing values)).
The regression tells me that the effect of DD coefficient, an interaction of Treatment and Time, is positive and significant (p-value is 0,0000000) and the effect is app. 20 % increase.
The model looks like this:
Code:
xtreg y Treatment##Time i.year, fe
I run some more regression and I saw that if I run the same regression without dummies accompanying DD coefficient (see code below), the DD coefficient is significant and negative.
Code:
xtreg y Treatment#Time, fe
Can you please help me how to visualize the data in a case of this DD or DDD regression so that I can accompany my work also with visualization?
Thank you very much
Vojtech
