Paneldata
Posted: Sat May 10, 2014 1:25 pm
Hello,
I'm new to Eviews and I have difficulties concerning the paneldata regression I want to make.
Although there is no huge correlation among my explaining variables, it seems that my outcome for the Fixed Effect Model suffers from multicollinearity (the R^2 value is quite high, although only 1 explaining variable is significant).
Only when I leave out the last explaining variable (capacityratio) I seem to get an acceptable outcome for the Fixed Effect Model. But still it is impossible to introduce any dummy variables concerning the time-effect (the error message 'Cannot find constant equivalent in equation specification. Unable to normalize equation effects estimates' comes on)
Does anybody know how to solve this?
Thank you
I'm new to Eviews and I have difficulties concerning the paneldata regression I want to make.
Although there is no huge correlation among my explaining variables, it seems that my outcome for the Fixed Effect Model suffers from multicollinearity (the R^2 value is quite high, although only 1 explaining variable is significant).
Only when I leave out the last explaining variable (capacityratio) I seem to get an acceptable outcome for the Fixed Effect Model. But still it is impossible to introduce any dummy variables concerning the time-effect (the error message 'Cannot find constant equivalent in equation specification. Unable to normalize equation effects estimates' comes on)
Does anybody know how to solve this?
Thank you