Zero inflated Independent Variables
Posted: Tue Oct 08, 2013 9:55 am
Hello
I would like your precious input to an estimation problem I have. I use a panel dataset of firms with same accounting variables where some variables are zero-inflated (in case that a firm does not have a certain observation in an accounting variable this variable is set to zero). Is it possible to use an ordinary LSDV estimator or I need another method of estimation that takes into account the zero inflated variables? What would be the consequences in the first case and what are the options for the second case. Thank you in advance.
Best
Kouvas
I would like your precious input to an estimation problem I have. I use a panel dataset of firms with same accounting variables where some variables are zero-inflated (in case that a firm does not have a certain observation in an accounting variable this variable is set to zero). Is it possible to use an ordinary LSDV estimator or I need another method of estimation that takes into account the zero inflated variables? What would be the consequences in the first case and what are the options for the second case. Thank you in advance.
Best
Kouvas