I'm working on my thesis and I need to derive the standard errors of impulse response function, not really sure how to do this. In specific, say, if my IRF is significant between periods 1 and 6, then how do I calculate the standard errors of my IRF in these 6 significant periods? I use EViews, EViews gives you corresponding standard errors of IRF at each t, making me confused - should I average these standard errors or add them up to get the accumulative standard errors for, say, the 6 significant periods?...Dont think we can add up standard errors...Hope someone can help? Thank you very much for your time!
http://www.eviews.com/help/helpintro.html#page/content%2FVAR-Views_and_Procs_of_a_VAR.html%23ww37080
Hi thanks. But I actually looked at those before and couldn't find relevant answer. I know how to let EViews give me standard errors. But my question is how to calculate standard errors for the whole significant period. Or, did I miss anything on the EViews instruction page?
The tabular representation gives standard errors for each period as taken in the study. Why do you need to add the SDs?
The cumulative standard deviation formula is derived from an SD formula called the Raw Score Formula, But is it needed in the study??
Hi,
Actually VAR is the first stage of my research, my second-stage regression will use the estimates of impulse response function as dependent variable, and I will weight it by corresponding standard errors, that's why I need to derive the standard errors for the whole significant period. I was wondering whether EViews just simply adds up the standard errors of each period to form the standard errors of the accumulated IRF...but now I think the standard error next to the accumulated IRF given by EViews should be correctly calculated rather than just simply adding up each of the non-accumulative errors. Thank you very much for your help!
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Writing