i am working on a paper related to 8 years of financial data of 631 banks in 4 countries. the question is when performing panel data do I create 631 variables and ignore the 4 countries or what? if so what is the difference if those 631 banks were related to one country or 4 countries?
All you need to do is introduce country-level variables into the mode alongside the bank variables. The country-level data/variables will control for country differences in the data. This is because the country data series will not be the same for each country.
Don't ignore the 4 countries. You should do the fixed effects for these countries.
Choosing the variables and whether to use one or more countries depend heavily on your study objectives, research question(s), and the problem you are planning to resolve. Using more than one country in your study may help you doing relevant comparisons. I suggest you read previous similar papers as guidance.
You will need to maintain the country identity in the data. Otherwise the analysis could become meaningless.
Emmanuel V Murray
you mean add control variable to control countries effect.
i tried panel analysis in STATA and by using xterg command. i do not if results are logical and if that means i will accept the random effect?
can you please guide me and check
- Badges
- Science topic
- Similar topics
- Biological Science
- Histology
- Sectioning