Page 1 of 1

confidence interval

PostPosted: Thu Apr 08, 2010 10:49 pm
by p.gelain
I have a doubt in interpreting the confidence interval in the dynare outcome. Isn't it improper to call them "confidence interval"? Isn't it correct to call them probability interval in the Bayesian context, given that they should correspond to percentiles? And if the latter case is correct, don’t we need another percentage other that 90% (90th percentile?) to indicate the lower percentile (10%, 10th percentile?)?

Thanks

Paolo

Re: confidence interval

PostPosted: Fri Apr 09, 2010 10:01 pm
by p.gelain
Ok, I found the solution by myselfe. Dynare provides deciles.

Paolo