Panel data and stationarity studies

[Zied]

Hi everybody,

I'm working on panel data (20 countries, 25 years) modelling destination competitiveness explained by several independant variables (FEModel). Given that I'm working on long time series I'll be grateful if you could answer these two questions (one econometric and the other is for eviews application):

1-Do the stationarity and cointegration tests matter in this case? if yes, what are the most suitable tests to run?
2- How to run these tests on Eviews (version 5.1)

Thanks in advance guys

[Zied]

could someone help me with this question?

[wangari]

hey, cant help with the eviews part ... yet, am hoping to understand more too on that part, but given that it is a panel with a long timeseries trend, yes, u do need to run certain diagnostics, and stationarity does matter too, depending on what tests u r running, and what u r looking for, panel cointegration might not be necessary...look up the different tests, POLS, REM, FEM and to determine which one best suits ur data set and answers ur questions, u will need to carry out either or both of Hausman or the Lagrange Multiplier (LM) tests...hope u get alittle better insight into what u want to achieve after you read up on these...

[wangari]

hehehe, and being new to this forum, i completely overlooked the dates part...eish, so here is hoping u did finally manage ur way around your data set, seeing as this reply is almost a yr late...