Dear Experts,
I am using panel data for running ARDL.
My question is that as I have same Rainfall quantity in many of my planting areas within Location Groups, these areas get stacked one over the other for use in Eviews as panel data.i.e.
Production Location Group Rainfall(mm)
2000 Area1 100
2001 Area1 120
2000 Area 2 100
2001 Area 2 120
Will this high correlation between rainfall areas (repeated rainfall figures) pose any problem in using ARDL ? If yes, what is the way around?
It will pose problems in using ARDL.
Two-step GLS estimation using the Prais-Winsten or Hildreth-Lu procedures can be a proper approach.
Good Luck.
yes, you use 2-step GLS estimation .
Use spatial analysis for your estimation
If I understand M. Azzam correctly, we are discussing a panel with two time periods in which the data for one of the explanatory variables is replicated across some of the x-section units. I don't see why this should prevent estimation in general. QUESTION: In the particular case of estimating ARDL, is there something about the preferred estimator(s) that could cause problems in the presence of this data feature? ANSWERS AWAITED
(By the way, ARDL with only two time-periods seems a very restricted version of ARDL.)
Really grateful to Hasret, Dinda and Daly. As for comments of Mr. Vince Daly, the data given by me above was for example. Actually I have 28 years of data. Yes I would like to see Hasret and Dinda resond to your query and mine query of why should such a repeated data caused problem.
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude