I have a panel data with 2700 firms and 8 years.
however, a lot of firms have missing data.
Is that will affect my result significantly ?
Is cleaning the dataset from outliers using Winsorising, will help to make the data better ?
How can I solve the issue for missing data ?
some articles mention that they clean the dataset from observation with missing value.
Is it necessarily to do that and how can I do it ?
e.g. is the code ?
drop variable if =.
Advertise on Talk Stats