Cumulative for frequent zero explanatory variable
Posted: Tue Apr 12, 2011 4:42 am
Dear Econometrician wizards 
, I want to consult with you a problem I faced in Econometrics.
Please reply to me as soon as you can, any idea would be appreciated.
I have a explanatory variable (advertisement expenditure) which is zero at so many times. It means from 160 obs, I have 74 obs as zero and 68 nonzero (positive values).
Now according to some research i read that the maximum 12 lags is suggested by industry experts. it means each positive observation has 12 lag effect .... now my question is:
if I calculate cumulative of advertisement which give me a step function, and the model is perfectly working.
Ramsey RESET is good, no seriall correlation and Rsq is 0.77, DW=2 and F p value o.ooooo ...
now my question is " can I do that"????
Thank you so much. Hope you won't stuck in any econometrics project ...
, I want to consult with you a problem I faced in Econometrics.Please reply to me as soon as you can, any idea would be appreciated.
I have a explanatory variable (advertisement expenditure) which is zero at so many times. It means from 160 obs, I have 74 obs as zero and 68 nonzero (positive values).
Now according to some research i read that the maximum 12 lags is suggested by industry experts. it means each positive observation has 12 lag effect .... now my question is:
if I calculate cumulative of advertisement which give me a step function, and the model is perfectly working.
Ramsey RESET is good, no seriall correlation and Rsq is 0.77, DW=2 and F p value o.ooooo ...
now my question is " can I do that"????
Thank you so much. Hope you won't stuck in any econometrics project ...