**Today:**

- Chapter 12 - Autocorrelation

**Next Time**:

- Continue Chapter 12 - Autocorrelation

**Video**:

« Material for Lecture 6 on Thursday 1/20/11 | Main | Homework 4 »

The comments to this entry are closed.

Mark Thoma

Department of Economics

University of Oregon 1285

Eugene, OR 97403-1285

Mark Thoma's Web Page

Email Mark Thoma

Mark Thoma's Blog

Office: PLC 471

Office Hours: T/Th 3:30-4:30

Gulcan Cil

Office: PLC 431

Email: gcil@uoregon.edu

Office Hours: M 1:00-3:00 and by appointment

Colin Corbett

Office: PLC 504

Email: corbett@uoregon.edu

Office Hours: T/Th 11-12 and by appointment

structure:

1.three models of serial correlation

if ut not correlated with any x, it is fine, unbiased

if ut only correlate with x2t in it own equation, biased but consistent

if ut not only correlate with x2t in it own equation, but also correlates with x from previous and after equations, biased inconsistent and inefficient( since inefficient problem, makes std error smaller than real, then make t-stat and f-stat very significant, everytime you see this big t, should consider time series problem)

2. Test for auto correlation

a.steps of durbin watson test(first order)

b.reject regions, inconclusive regions, and fail to reject regions of the D-W test.

c. prove 0

Posted by: Ji | Wednesday, August 08, 2012 at 07:07 PM