Karolina, It sounds like your questions would be better addressed by consulting an econometrics textbook (and of course the gretl user's guide). This list is really meant for questions like "can gretl do xxx test for panel data?" Rather than "what tests shouldI do on panel data?" PS ________________________________________ From: gretl-users-bounces(a)lists.wfu.edu [gretl-users-bounces(a)lists.wfu.edu] on behalf of Karolina Zinkeviciute [karolina.zinkeviciute(a)gmail.com] Sent: Tuesday, April 24, 2012 6:56 PM To: gretl-users(a)lists.wfu.edu Subject: [Gretl-users] help with panel data Hello gretl users:) I am asking for some help with panel data. It's the first time that I'm working with this type of data. So I'm not fully aware of the steps I should make. I don't know what should I do before proceeding to the fixed effects test. What tests should I use for Heteroskedasticity, Autocorrelation and serrial correlation for panel data? What other test should I make? I will be very grateful for any hints:) Karolina

Hi Karoline                  As far as I know: -Ramsey RESET is not for Panel OLS -ADF is for time-series variables: check on first and second generational panel unit root tests(LLC, IPS,....etc) for panel unit root tests -the purpose of Jacque Berra is almost same as that of panel unit tests              As Summer Peter just adviced, take sometime and digest Gretl User's guide. Panel regression is multidimensional . Gretl User's guide will show how to: distinguish between FE and RE regressions, how to explore dynamic panels, how to use Instrumental panel regressions.....etc. ________________________________ From: Karolina Zinkeviciute <karolina.zinkeviciute(a)gmail.com&gt; To: Gretl list <gretl-users(a)lists.wfu.edu&gt; Sent: Wednesday, April 25, 2012 8:09 AM Subject: Re: [Gretl-users] help with panel data Thank you for your remark.  I realize have I put the question in the wrong way. For sure I have consulted textbooks and gretl user's guide before writing here. Though, I could not find whether I can use the same tests for panel data as I use with time-series or cross-sectional data So I would like to ask if it is ok to use these tests for panel data:     white test for testing heteroskedasticity     Augmented-Dickey fuller test for serrial correlation,     Ramsey RESET test for correct model specification,      and jacque berra test for nomality? I would highly appreciate an advise as I was not able to figure that out myself. Karolina On 25 April 2012 02:23, Summers, Peter <psummers(a)highpoint.edu&gt; wrote: Karolina, -- Karolina Zinkevičiūtė _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-users