LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous (more recent) messageNext (less recent) messagePrevious (more recent) in topicNext (less recent) in topicPrevious (more recent) by same authorNext (less recent) by same authorPrevious page (November 2004, week 3)Back to main SAS-L pageJoin or leave SAS-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:         Tue, 16 Nov 2004 11:59:32 -0800
Reply-To:     Tracy Li <lisiqi77@YAHOO.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Tracy Li <lisiqi77@YAHOO.COM>
Subject:      tests of fixed effects
Content-Type: text/plain; charset=ISO-8859-1


I am running an unbalanced panel data analysis (100 countries in 10 years) by using proc mixed (I am treating country as random and year as fixed):

proc mixed data=a; class country year; model y=year x1 x2 x3; random country; run;

The output contains a type 3 tests of fixed effects like this, effect pr>F year <0.0001 x1 0.05 x2 0.99 x3 <0.001

Now I am interested in finding out if it is appropriate to treate year as fixed and country as random. So according to the table above, can I just say I reject the null of fixed effects for year--i.e., I should treat year as random too?

PS, when the codes are like:

proc mixed data=a; class country year; model y= x1 x2 x3; random country year; run;

The tests for fixed effects are <0.0001 for both year and country.

How should I read those test results, what is the null hypothesis here?

Many thanks, Tracy Li

Back to: Top of message | Previous page | Main SAS-L page