I have a gravity model with a panel data But when I begin to estimate my model with all of my variables, there was an e-views error stated "insufficient number of observation". I try to put only several variables on my model, then it works (but only works with Panel Least Square Estimation). I still can't put all of my variables on my model Do you know why I can't running an estimation with all of my variables there? Does gravity model have a limitation for its variable specification? What should I do?
plz explain how much you have total observations, cross sections , independent variables AND want to include time fixed effect or not
Dear Anandita,
There is a relation between observation number and indipendent variable numbers. While the number of indipendent variables is incresing, the level of freedoom decreses. Or while number of observation is decreasing, the level of freedoom decreases as well. So some statics cannot be estimated. In this case you should either increase the number of observation or decrease the number of indipendent variables.
Dear Anandita I would like to know what is your model for helping you. Is it possible to send me it.
you should decrease independent variables so that degrees of freedom should increase.
Ok but first is important define the model and consider the variables requiered.
Same with me the error message shows when running the data for Unit Root test at 1st difference.
- Badges
- Science topic