Page 1 of 1

Significant Difference

PostPosted: Sun Nov 11, 2012 4:16 am
by samster
Hi:

Suppose I adjust a model (maybe change a parameter value) and my output changes(e.g. IRFs indicate a bigger or smaller response), is there a way to calculate if said changes are statistically different from the "benchmark" situation?

Thanks.

Re: Significant Difference

PostPosted: Mon Jan 07, 2013 1:17 pm
by SébastienVillemot
You may want to have a look at the sensitivity analysis embedded in Dynare, thanks to Marco Ratto:

http://www.dynare.org/manual/index_27.h ... n-analysis
Ratto, Marco (2008): “Analysing DSGE models with global sensitivity analysis”, Computational Economics, 31, 115–139

Re: Significant Difference

PostPosted: Fri Jun 07, 2013 9:48 pm
by samster
Thanks.