Page 1 of 1

Panel data, fixed effects

Posted: Fri Sep 12, 2014 10:04 am
by daan
Hello,

I'm using panel data and have two dummies, dfinancial (when the firm is a financial firm) and dcrisis(wheter it is a crisis year).
When i'm trying to figure out if i should use fixed of random effects i get stuck.

I think i can't use a firm specific dummy when using fixed firm effects and i can't use period fixed effects when using time dummies.

Is it correct to say i should use random effects? The hausman test say's i should use fixed effects.

Or should is split up my data set in two sets, on financials on other and use a chow test in order to test if the coefficients are sign. different?

please help me!!