Page 1 of 1

Redundant Variables Test

Posted: Mon Feb 20, 2012 6:39 am
by rushpinn
Hi!

We got an assignment where we are suppose to make a multiple regression analysis. I estimated the regression with LS (4 independent variables) and then we are asked to perform a "redundant variables test". Now, I wonder: Since I can already see by the p-values in the estimation output which variables are significant, is there any meaning/advantage in doing a joint test like the redundant variables test? or is our professor just trying to teach us HOW to make one. So that we will remember to make one when we get to e.g. panel data or times series assignments?

Re: Redundant Variables Test

Posted: Mon Feb 20, 2012 7:21 am
by startz
There is a good reason. The answer you get from a joint test is not the same as the answer you get from a series of tests on individual coefficients. The joint test is the right way to test a joint hypothesis.