First - Professor Sekhon emailed to say that " conditioning
on the fact that his hip has just been replaced, he feels fine." That
manner of speaking suggests to me that his brain if up and running as
usual.
A few items:
(1) Homework Seven, Second Solution Set has been posted to the
course website, as of early last night. You will want to look over this
as you prepare. Probably it will be more useful to you as you actually
work on the final, as I have left a number of hints and directions
throughout. If anything is puzzling about that document, let me know.
(2) Email during the final. Email me directly with any substantive
question so that I can edit out any content that 'gives too much away.'
If you get stuck on any aspect of the coding, email the list directly.
I will check my email at the following times 7 a.m. 10 a.m. 2 p.m. 5
p.m. 8 p.m. 11 p.m. So expect responses accordingly.
(3) Turning in the final. What you turn in must be in following
form. Two items: (1) A write-up of the final, not containing any R
code. This may be either turned into my box or emailed to me. If it is
emailed as a Microsoft Word document, the file name of this document
should be like 'klinefinalexam.doc.' (2) A file containing the R code
which you used to produce your results. I should be able to cut and
paste this file into R and reproduce your results. This file must be
emailed to me and should have a filename like 'klinefinalRcode'.
(4) Timing. The following students have requested not to receive
the final until Sunday, so as to have more time to study in groups:
bodnar, Mohandas. The files for the final will be on the course website
at 12:01 a.m. Friday morning. You will have to log in with your student
id and pin number - this will allow me to keep track of who gets the
final at what time. If you have any trouble doing that, let me know.
The final is due on Friday, January 16th.
e
(5) Office Hours. I will hold office hours tonight, from 10 p.m. to
12 p.m. in the lounge, basically to answer any last minute questions you
may have discovered. Feel free to drop by.
Finally, if you have any trouble, do not be afraid to email.
Good luck,
Jacob
Thank you for your responses. The meeting will be held on Wednesday,
January 21 at 3:30 pm.
>Date: Fri, 19 Dec 2003 15:09:16 -0500
>To: gov1000-list(a)fas.harvard.edu
>From: Kim Schader <kschader(a)latte.harvard.edu>
>Subject: Meeting to Discuss Methods
>Cc: kquinn(a)latte.harvard.edu, apost(a)fas.harvard.edu, king(a)harvard.edu
>
>
>Prof. Gary King asked me to write to each of you to set up a meeting, the
>subject of which is the methods sequence, the role of methods in political
>science, and any other big picture issues that might be of interest to you.
>
>Can you please tell me via email (kschader(a)latte.harvard.edu) which of the
>following meeting times you would be able attend? If you are able to
>attend more than one, tell me which would be your first, second, etc
>choice. I will pick the time that works best for the most people.
>
>The options are --
>
>Tuesday, January 20: 10:00 am - 11:00 am
>Tuesday, January 20: 2:30 pm - 3:30 pm
>Wednesday, January 21: 9:30 am - 10:30 am
>Wednesday, January 21: 3:30 pm - 4:30 pm
>Thursday, January 22: 4:00 pm - 5:00 pm
>
>Thank you in advance for your cooperation!
>
>Regards,
>Kim
Kim Schader, Project Coordinator
Center for Basic Research in the Social Sciences
Harvard University 34 Kirkland Street Cambridge, MA 02138
Phone: 617.495.9271 Fax: 617.384.7489
kschader(a)latte.harvard.edu
-----Original Message-----
From: Alex Liebman [mailto:liebman@fas.harvard.edu]
Sent: Tuesday, January 06, 2004 2:31 PM
To: gov1000(a)fas.harvard.edu
Subject: R Code questions
Hey Jacob,
Hope you a great break.
I had a couple questions about the R Code you sent out.
1) I couldn't figure out which F Test results corresponded to which line
in
the table. For example, for Model 2, which F Test produces the number
for
"sign. infl., chunrate."?
So sign. infl., chunrate. Goes with ftest(model1,comptest) because is it
the comparison between model1 and the least common model, which I've
called comptest.
Also, when I run your code I'm getting different
numbers.
OK then, either my numbers are wrong, my code is wrong, your code is
wrong, my data is wrong, or your data is wrong.
If you want a better response, you might send whatever R code you have
showing what you are getting wrong and then I can look at it.
2) In your code, why does the Vietnam variable get progressively more
positive? Shouldn't the effect of Vietnam get progressively more
negative?
It doesn't matter whether it gets progressively more positive or
negative. If my variable gets progressively more positive, then it will
have a negative coefficient. If a different variable got progressively
more negative, then the coefficient would be positive.
Thanks,
Alex
I have an emailed, MS Word document for hw7, part two, which is title
"Homework Seven part deux." But it does not contain a name. Anyone
want to claim it?
-----Original Message-----
From: Paul Bodnar [mailto:bodnar@fas.harvard.edu]
Sent: Sunday, January 04, 2004 4:18 PM
To: jkline(a)fas.harvard.edu
Subject: Three questions
Hi Jacob,
I hope you had a good break.
I have three short questions related to the final.
1) Will the math questions cover the entire semester or just the second
half?
The entire semester.
2) Can we expect the replication be a variation of the presidential
approval
data set (as Jas seemed to suggest) or an entirely different data set?
It is basically the same data set, but with some modifications to force
you to make some changes to the other assignments.
3) As my study group will not be able to get together until Wednesday, I
was
wondering if the following arrangement would work: the final is handed
out on Friday the 9th. Could my group agree not to start (or even take
delivery of) the final until, say, Monday the 11th so that we could have
more time to
prepare together? Obviously we'd still turn it in on the 16th, but I
think
we'd be willing to forgo some exam time to gain more prep time.
OK - To accommodate these kinds of requests, I will password protect the
folder for the final exam. You are on your honor, of course, not to
share the password with one another. Let me know when you would like
the password mailed to you. If you DO NOT RESPOND to this email, I will
email you the password at 12:01 a.m. Friday morning.
Also, is there any news on Jas' operation?
To be honest, I have no news. Jas is hopefully recovering in peace --
and in peace means 'away from his computer.' I will let you know what I
hear.
Thanks,
Paul
--
Paul Bodnar
Department of Government
Harvard University
+1 617 230 4525
bodnar(a)fas.harvard.edu
I will respond later tonight about the qq plots. I want to look over
the final first and see what one needs to know and then find a place in
the textbooks.
As for the command 'as.logical" : I use this command to convert a
series of ones and zero into FALSE and TRUE. Sometime the input for a
particular command requires that you have a vector of logical values
rather than binary integers. Alternatively, if you need to take TRUE
and FALSE to the integers 0 and 1, you can use the command 'as.integer'
So for example, let's say that you want to know how many of your z
scores are greater than 2.
zscores <- c(2.12,.14,.16,-4.99,-.4)
greaterthantwo <- abs(zscores) > 2
# Now does the vector 'greaterthantwo' have the values 0 and 1 or FALSE
and TRUE?
# Probably - logical values; but we can be more certain.
# Look at
as.logical(greaterthantwo)
as.integer(greaterthantwo)
-----Original Message-----
From: Jason Lakin [mailto:jlakin@fas.harvard.edu]
Sent: Monday, January 05, 2004 3:43 PM
To: gov1000(a)fas.harvard.edu
Subject: RE: qq
hi jacob. happy new year. i hope you had a good break. i am back in
town trying to get my act together for this final. i need help with qq
plots and the like. where are the notes about this? i don't see
anything in the lecture notes, and don't know where else to look. were
there some handouts that you can put on the website? are qq plots
covered in any textbook?
also---i don't understand how you used the as.logical term in your lm
code from the problem set. what does the command do?
thanks.
all the best
jason