ECON 710 Midterm: ECON 710 UW Madison Midterm 2009a
Document Summary
Since the two regressor matrices span the same linear space, the two residual vectors are identical, so the variance estimates are identical. 0 which is a full rank matrix. regression residuals are (it is upper diagonal with positive diagonal elements. 2. (a) q = 5: the dimension of (cid:12) 2 is 5, and the hypothesis is that all components of (cid:12) 2 are zero, which is (cid:133)ve restrictions. (b) no. The wald test rejects for large values of wn, when wn (cid:21) c for some c: a test which rejects for small wn can have correct type i error, but will have low power. The question about the 1% quantile is misleading. It is also not sensible to talk about test with 99% type i error. So there is no sense in which 0. 55 is a reasonable critical value for a test: the model is a linear regression.