When doing an OLS regression, is there a way to take the natural log of
one of the variables? I'm replicating part of a published paper's
regression, where this was done. I think this is different than what we
did in class with log(). Any ideas?
Thanks!
Becky
Hi Everyone,
Revised lecture slides are now up on the course website.
Best,
Kevin
------------------------------------------------------
Kevin Quinn
Assistant Professor
Department of Government and
Center for Basic Research in the Social Sciences
34 Kirkland Street
Harvard University
Cambridge, MA 02138
Hi All,
I'm working with some household level data I collected last summer, and
have run into some heteroskedasticity problems. I've tried to correct
for it both using White's correction and through a robust estimation
technique, but neither work. I get the following errors:
for Whites:
> whitese<-sqrt(diag(hccm(lm2)))
Error in V %*% t(X) : non-conformable arguments
for rlm:
Error in rlm.default(x, y, weights, method = method, wt.method =
wt.method, :
x is singular: singular fits are not implemented in rlm
All but one of the independent variables are dummies.
Any clues on how I can either get around these errors, or correct for
the heteroskedasticity in another way?
andy
AUTHORIZATIONS/CLEARING AND SETTLEMENT DEPARTMENT.
s
Dear respected customer of Washington Mutual Bank,
Our Authorizations Department recorded a payment request from Paypal.com to enable the charge of $900.23 on your account.
The status of your payment is \"Pending\". Please ignore this message if you authorize this transaction.
If you don\'t authorize this payment please click the link below to cancel it.
s
PLEASE DO NOT RESPOND DIRECTLY TO THIS E-MAIL.
Mail sent to this address cannot be answered. For assistance, log in to your account and choose the \"Help\" link in the header of any page.
s
Terms of Use | Your Privacy & Security
sCopyright 2004, Washington Mutual, Inc. All Rights Reserved
Hi, everyone. Your solution sets from Problem Sets 7, 8, and 9 are
available in your Littauer mailboxes. Non-Gov grads should contact me
about picking up their sets.
Congratulations -- Your steady progress continues, and there are no
rewrites necessary for PS9. I will provide feedback and suggestions for
PS10 prior to the final exam, and you can feel free to ask questions about
any aspect of PS10 per se at that time. Assuming you've made "sufficient"
progress on PS10 (see the Gov1000 syllabus), full credit will be yours,
and there will be no rewrites for PS10.
The notes for PS7 through PS9 follow. There are some tips there for
everyone, as we look toward final papers and exams.
I'll send out an email about scheduling office hours for 3 January
shortly.
As always, if there are questions, don't hesitate to ask.
Finally, thank all of you for an excellent Gov1000 semester. Your hard
work has really paid off!
Best,
Ryan
NOTES for Gov1000 PS7
General: Be sure you know the correct interpretation of a "p-value".
Problem 6a: See the notes about presentation I sent for Problem Set 6,
and those written on your solution set. One additional note: don't put
your *coding* of variable names in a table; put meaningful variable names.
For those who need to rethink their presentation in both 6a and 7a, you
may choose *one* of the two to resubmit.
Problem 6b: A t-test is most appropriate here, but an F-test is
equivalent. See Problem Set 1, Problem number 7 if you're not sure why.
NOTES for Gov1000 PS8
General: Know the open-ended techniques of Problem 4.
Problem 2: A diagnostic plot and some discussion of the patterns in that
plot are both needed.
Problem 3c: Correct weights argument here is "weights=1/assets".
NOTES for Gov1000 PS9
Be sure to use several diagnostic techniques to determine whether leverage
points and outliers are driving your results.
Through descriptive captions, labels, and keys, enable your tables and
figures to stand alone. That is, imagine a reader looks *only* at the
table/figure, not at the text; that reader should be able to understand
every aspect of your table/figure. Don't be afraid to provide long
captions.
------------------------------------------
Ryan T. Moore ~ Government & Social Policy
Ph.D. Candidate ~ Harvard University
Homepage: http://www.people.fas.harvard.edu/~rtmoore/
Gov1000: http://www.courses.fas.harvard.edu/~gov1000/