Fixed vs. Random Effects Models. Hausman test.

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

ryan_DETROIT
Posts: 12
Joined: Wed Mar 31, 2010 5:19 am

Fixed vs. Random Effects Models. Hausman test.

Postby ryan_DETROIT » Wed Mar 31, 2010 6:07 am

hi all,
i am trying to select which models i should use to conduct panel data analysis. random vs. fixed effects.
In order to make a choice between random effects model and fixed effects model i should perform Hausman test. Yet i do not know how to perform Hausman test. does anybody help me ? I can not find the hausman test button on EViews program.
many thanks.

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13586
Joined: Tue Sep 16, 2008 5:38 pm

Re: Fixed vs. Random Effects Models. Hausman test.

Postby EViews Gareth » Wed Mar 31, 2010 6:48 am

Estimate an equation with random effects, then click on View->Fixed/Random Effects Testing->Hausman Test.

It is fairly well documented in the manual/help system.

ryan_DETROIT
Posts: 12
Joined: Wed Mar 31, 2010 5:19 am

Re: Fixed vs. Random Effects Models. Hausman test.

Postby ryan_DETROIT » Wed Mar 31, 2010 6:53 am

thanks for the reply. i am using EViews 5.1 version.
there is no button view> Fixed/Random Effects Testing->Hausman Test.
how i can estimate an equation with random effects ? Do you mean write an equation on the command bar ?

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13586
Joined: Tue Sep 16, 2008 5:38 pm

Re: Fixed vs. Random Effects Models. Hausman test.

Postby EViews Gareth » Wed Mar 31, 2010 7:06 am

Perhaps you would be best off reading the panel estimation chapter of the help/manual, I think it explains it all pretty well.

ryan_DETROIT
Posts: 12
Joined: Wed Mar 31, 2010 5:19 am

Re: Fixed vs. Random Effects Models. Hausman test.

Postby ryan_DETROIT » Wed Mar 31, 2010 1:58 pm

i have read EViews 5.1 guide. yet it did not help me to solve my problem. I just wanna know how to run hausman test in order to select method. does anybody help me please ? by the way yearly five independent data and one dependent data between 1993 and 2008 is sufficient for panel data analysis?
many thanks...

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13586
Joined: Tue Sep 16, 2008 5:38 pm

Re: Fixed vs. Random Effects Models. Hausman test.

Postby EViews Gareth » Wed Mar 31, 2010 2:24 pm

hausman2.png
hausman2.png (296.64 KiB) Viewed 27638 times
hausman.png
hausman.png (112.94 KiB) Viewed 27637 times

ryan_DETROIT
Posts: 12
Joined: Wed Mar 31, 2010 5:19 am

Re: Fixed vs. Random Effects Models. Hausman test.

Postby ryan_DETROIT » Thu Apr 01, 2010 2:11 am

that buttons do not appear in EViews 5.1 which i use. i want to have result spreadsheet below.

Hausman test

(fixed versus random effects)
Chi-square (2 d.f.) 8.2380468
p-value 0.0162604

ryan_DETROIT
Posts: 12
Joined: Wed Mar 31, 2010 5:19 am

Re: Fixed vs. Random Effects Models. Hausman test.

Postby ryan_DETROIT » Thu Apr 01, 2010 2:07 pm

hi there, i just want to run Hausman test in order to select method ( Random effects or fixed effects). \
in Stata ,researchers are able to run Hausman test by comparing coefficients of Fixed or Random effects. is it not available in EViews 5.1 ?
thank you...

EViews Glenn
EViews Developer
Posts: 2682
Joined: Wed Oct 15, 2008 9:17 am

Re: Fixed vs. Random Effects Models. Hausman test.

Postby EViews Glenn » Thu Apr 01, 2010 4:01 pm

We are still puzzled as to why this isn't working for you since it is certainly a part of EViews 5.1.

Let me make certain that you...

1. Have a panel structured workfile or a pool object in a non-panel workfile.
2. Have estimated your equation or pool using random effects.
3. Are looking on the View menu of the estimated equation or pool.

Note that this feature was not available in EViews 5 so that EViews 5.1 is required.

If you believe that the answer to all of these questions is yes, you probably should send us your workfile, along with the build date of your copy of EViews. You can find the latter by selecting Help/About EViews from the main menu.

ryan_DETROIT
Posts: 12
Joined: Wed Mar 31, 2010 5:19 am

Re: Fixed vs. Random Effects Models. Hausman test.

Postby ryan_DETROIT » Fri Apr 02, 2010 5:33 am

ok, i agree with you, i should probably send my workfile in Excel format. maybe now you can help me...many thanks...
Attachments
Kitap1.new.START.xls
Excel fie.for EViews analyze.
(33 KiB) Downloaded 765 times

EViews Glenn
EViews Developer
Posts: 2682
Joined: Wed Oct 15, 2008 9:17 am

Re: Fixed vs. Random Effects Models. Hausman test.

Postby EViews Glenn » Fri Apr 02, 2010 10:21 am

You should send the *EViews workfile* along with your serial number to tech support (support@eviews.com) and we'll be happy to take a look.

ryan_DETROIT
Posts: 12
Joined: Wed Mar 31, 2010 5:19 am

Re: Fixed vs. Random Effects Models. Hausman test.

Postby ryan_DETROIT » Tue Apr 06, 2010 10:29 am

sorry for the late answer. I am using EViews 5.1. But i can not find a button that enables me to run Hausman test. I need to run Hausman test in order to make choice between random effects model or fixed effects model. BTW pic. was too large. then i zipped that .
Attachments
EViews.jpeg.rar
(71.96 KiB) Downloaded 893 times

EViews Glenn
EViews Developer
Posts: 2682
Joined: Wed Oct 15, 2008 9:17 am

Re: Fixed vs. Random Effects Models. Hausman test.

Postby EViews Glenn » Tue Apr 06, 2010 11:29 am

Please send the *EViews workfile* along with your serial number to tech support (support@eviews.com) and we'll be happy to take a look. You may zip the workfile if you like. Please don't it send in RAR format.


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 2 guests