Res Ec 703: Econometrics
Order
of
Topics And Readings For Lecture 17, November 4, 2010
Source
Exam 2
Assignment
6
Note
BM
Bottom 85
SAS
1431
BM
9
Kennedy
130
Monday, November
22,2010;
Stockbridge 114;
6:00 p.m. to 8:00 p.m.
Due Tuesday, November
9,2010
We
meet next Wednesday but not next Thursday.
(Wednesday, Nov. 10 follows Thursday, Nov.
11
's
schedule).
Material Distributed Last Class But Not Covered
White's
polynomial expansion specification as the form for
heteroscedasticity. The steps in White's test are summarized
at the bottom
ofBM
page 85.
Does SAS provide assistance with getting this estimator?
Testing For Heteroscedasticity (The Reg Procedure):
If
you suspect heteroscedasticity, get the "nice", i.e., the
consistent, estimators
of
the standard errors
of
the regression
coefficients. The ACOV option is the answer.
This page (specifically, equation 16) is revisited because it
provides justification for what is done on SAS p. 1431.
Kennedy's blurb about White along with
BM's
"long hand"
calculation
of
consistent standard errors. Compare these to
the ACOV output.
End
Of
Material Distributed Last Class But Not Covered
GHJ
Gujarati
BM
BM
Chapter 16
222
86
Revised
BMPage
87
Autocorrelation
How autocorrelation surfaces when a model is incorrectly
specified: autocorrelation as the result
of
misspecijication.
(This is also called induced autocorrelation).
Dealing with autocorrelation even after everything is done to
specify a model correctly.
CochraneOrcutt methods (Substitute this page for current
BM
page 87.)
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Basic Econometrics by
Damodar
Gujarati
(First
edition,
page
222)
One
possible
reason
for
the
presence
of
autocorrelation
2.
This is the end of the preview.
Sign up
to
access the rest of the document.
 Fall '10
 BernardMorzuch
 Statistics, Econometrics, Regression Analysis, Autocorrelation, strong positive autocorrelation, strong negative autocorrelation

Click to edit the document details