Thread: Summing Regressors in a regression

Can you sum the co-efficients on regressors to get a better approximation of something? And then get the 95% confidence interval?

If so how would you do this?

To me this would be a problem as you can't sum standard errors which are necessary for getting the confidence interval? So you would need the standard errors of the sum of the regressors???

Thanks in advance

Hey froggy124.

Can you give an example of this please symbolically?