Testing significance of results- please help!
Posted: Thu Mar 11, 2010 11:54 pm
Hello,
I am writing my senior econometrics thesis, and have a problem.
I want to test whether 7 values that I forecasted for credit spreads of a bank is SIGNIFICANTLY different from the actual credit spreads that were observed. Is there a way I can test for this significance? I was wondering if a t-test would suffice.
To reiterate:
12/1 forecast 1
12/2 forecast 2
....
12/7 forecast 7
OBSERVED Values:
12/1 observation 1
12/2 observation 2
....
12/7 observation 7
How can I test whether the difference (residuals) between the forecasted and observed are significant? Can I do such a test in eviews?
Thanks so much in advance.
I am writing my senior econometrics thesis, and have a problem.
I want to test whether 7 values that I forecasted for credit spreads of a bank is SIGNIFICANTLY different from the actual credit spreads that were observed. Is there a way I can test for this significance? I was wondering if a t-test would suffice.
To reiterate:
12/1 forecast 1
12/2 forecast 2
....
12/7 forecast 7
OBSERVED Values:
12/1 observation 1
12/2 observation 2
....
12/7 observation 7
How can I test whether the difference (residuals) between the forecasted and observed are significant? Can I do such a test in eviews?
Thanks so much in advance.