I am a new user of panel regression. I am trying to explore the determinants of FDI at a global scale. I have data of 124 countries for 19 years where some data for some countries on FDI itself for some years are missing. Alongside, I have also missing values for my every independent variables for some years (i.e. for some variables in beginning years, for some in middle years).
In this situation I need help to solve few questions:
1. Can I keep the missing values as they are and run the regression and write the paper? Up to how much (i.e. in % of total data) missing value is keepable in a panel regression?
2. I want to use ipolate and epolate for estimating missing values. Is it alright? Can I use these for dependent variable too?
3. Is missing value in dependent variable acceptable? or shall I remove the countries or cases for missing dependent variables?
4. How shall I determine which will be my X var in using ipolate and epolate?
Would be very happy if anyone can suggest me a way.
In this situation I need help to solve few questions:
1. Can I keep the missing values as they are and run the regression and write the paper? Up to how much (i.e. in % of total data) missing value is keepable in a panel regression?
2. I want to use ipolate and epolate for estimating missing values. Is it alright? Can I use these for dependent variable too?
3. Is missing value in dependent variable acceptable? or shall I remove the countries or cases for missing dependent variables?
4. How shall I determine which will be my X var in using ipolate and epolate?
Would be very happy if anyone can suggest me a way.