random effects-help pls
Posted: Sun Jan 15, 2012 12:55 am
Dear friends,
I am working on panel data model.(18 countries and 2005-2010)
Unfortunately,I am confused about using Coef covariance method.
I mean when I use white cross-section method in Coef covariance method :
p - value in Hausman test will be :1
(I revived a warning message about evies set p vlaue on zero)
when I use white period method in Coef covariance method :
p - value in Hausman test will be :0
when I use cross section sur method in Coef covariance method :
p - value in Hausman test will be :0.33
In this condition,fixed effects should be applied or random effects???
Moreover, R2 is nearly 0.29 in random effects approach.Is not it very low?
Help me please.
Tanks in advance.
Amir
I am working on panel data model.(18 countries and 2005-2010)
Unfortunately,I am confused about using Coef covariance method.
I mean when I use white cross-section method in Coef covariance method :
p - value in Hausman test will be :1
(I revived a warning message about evies set p vlaue on zero)
when I use white period method in Coef covariance method :
p - value in Hausman test will be :0
when I use cross section sur method in Coef covariance method :
p - value in Hausman test will be :0.33
In this condition,fixed effects should be applied or random effects???
Moreover, R2 is nearly 0.29 in random effects approach.Is not it very low?
Help me please.
Tanks in advance.
Amir