Dear all,
I'm working on a panel data for 18 countries during the period 1960-2016. I ran Johansen-Fisher Cointegration test. According to trace statistics and Max Eigen statistics, there exists one cointegrating relationship between the variables.
However, most of the countries are dropped in the "individual cross sections results". I ended up with 4 countries out of 18 as attached in the picture below. Could anybody tell me the reason ? and how I can fix this? I believe it's not related to any missing values as I've used linear interpolation for missing values.
Any advice will be highly appreciated.
Thanks in advance.
Dina
First, I would advise you to look descriptive statistics. In STATA you can check the descriptive stat with xtsummary commamd. That could indicate if you have some scaling problems, or some general data issue. Secodn, have you ploted the data?
There should be an issue with your data. As mentionned before the fisrt step would be to look to your graphs and desc. stats.
Dear Bojan BASKOT & Mohamed DOUCH,
I checked the descriptive statistics of my data as attached in the picture below. General remark before interpolation : lots of missing values and zeros. What do you think?
Thanks for your reply.
- Badges
- Science topic
- Similar topics
- Philosophy
- Philosophy Of Science
- Reasoning