But avar n n avar n n g c 2 g g c 1 g g c 2 c 1 g and

Info iconThis preview shows pages 76–77. Sign up to view the full content.

View Full Document Right Arrow Icon
But Avar n ̃ n  Avar n ̂ n  G C 2 G G C 1 G G  C 2 C 1  G and this is a psd matrix because C 2 C 1 is psd . 76
Background image of page 76

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
The conclusion here is important and not surprising: If we have an asymptotically efficient estimator of , we should use it in constructing an estimator of g . The fact that smooth, nonlinear functions of are better estimated using an asymptotically more efficient estimator further illustrates the flexibility of asymptotic analysis. In finite-sample analysis, we can only be sure that linear functions of a parameter are better estimated (have smaller variance) using a best unbiased estimator. 77
Background image of page 77
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

Page76 / 77

But Avar n n Avar n n G C 2 G G C 1 G G C 2 C 1 G and this...

This preview shows document pages 76 - 77. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online