Statistics for polled model with random effect

kiber_master · Postby **kiber_master** » Tue Dec 13, 2011 4:37 am

Hello!
Could you explained, why Akaike info criterion, Schwarz criterion and Hannan-Quinn criterion don't calculating for polled model with random effect?

EViews Glenn · Postby **EViews Glenn** » Tue Dec 13, 2011 10:16 am

Estimation of this model in EViews is not performed by maximizing a likelihood. Thus, we do not report a likelihood nor the related IC statistics.

kiber_master · Postby **kiber_master** » Tue Dec 13, 2011 10:54 pm

For model without effects (or for OLS model) they are calculated, why don't you show them for model with random effect? OLS model definitely don't need maximizing a likelihood.

EViews Gareth · Postby **EViews Gareth** » Tue Dec 13, 2011 11:40 pm

OLS is maximum likelihood.

EViews.com