We are modelling the effect of the number of ballot propositions on voter
turnout. Scholars of ballot propositions have traditionally used OLS
models to estimate this effect, but this seems implausible. It seems
likely that there will be diminishing returns (a/k/a "voter fatigue"), so
that the model should not be linear. For instance, the difference in
turnout for 3 ballot propositions vs. 0 ballot propositions is probably
not equal to the difference in turnout for 6 ballot propositions vs. 3
ballot propositions. How can we estimate these threshholds, or effectly
model this nonlinearly?
I recognize this is a problem of estimating threshholds for an ordered
categorical variable, but I am not sure how to approach it because the
categorical variable at issue is our *explanatory* variable, not our
dependent variable. Any suggestions?
Thanks,
Anna
This is a multi-part message in MIME format.
------=_NextPart_000_000D_01C31232.7080CD60
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Hi Everybody
I have five imputed data sets that i got from running Amelia. Relogit =
works fine, so does clarify and Amelia did too. I am having difficulty =
running clarify on the logit command in Stata and also would like to =
know how I can use the imputed data sets with relogit, since relogit is =
not supported by clarify (according to the message I got from stata.) =
The problem seems to happen when i use the mi(filename filename . . . ) =
part after estsimp logit y x1 x2 x3 . . . . When i put in the mi() part =
I get a message "c ambiguous abbreviation." Has anybody run into this =
yet?
T
------=_NextPart_000_000D_01C31232.7080CD60
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Diso-8859-1">
<META content=3D"MSHTML 6.00.2719.2200" name=3DGENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=3D#ffffff>
<DIV><FONT face=3DArial size=3D2>Hi Everybody</FONT></DIV>
<DIV><FONT face=3DArial size=3D2></FONT> </DIV>
<DIV><FONT face=3DArial size=3D2>I have five imputed data sets that i =
got from=20
running Amelia. Relogit works fine, so does clarify and Amelia did =
too. I am having difficulty running clarify on the logit command =
in Stata=20
and also would like to know how I can use the imputed data sets with =
relogit,=20
since relogit is not supported by clarify (according to the message I =
got from=20
stata.) The problem seems to happen when i use the mi(filename =
filename .=20
. . ) part after estsimp logit y x1 x2 x3 . . . . When i put in the mi() =
part I=20
get a message "c ambiguous abbreviation." Has anybody run into =
this=20
yet?</FONT></DIV>
<DIV><FONT face=3DArial size=3D2></FONT> </DIV>
<DIV><FONT face=3DArial size=3D2>T</FONT></DIV></BODY></HTML>
------=_NextPart_000_000D_01C31232.7080CD60--
This is a multi-part message in MIME format.
------=_NextPart_000_0020_01C31238.0139A750
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Okay, I got clarify to run on just plain logit using Amelia. But is =
there some way to use amelia data sets with relogit?
Traci
------=_NextPart_000_0020_01C31238.0139A750
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Diso-8859-1">
<META content=3D"MSHTML 6.00.2719.2200" name=3DGENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=3D#ffffff>
<DIV><FONT face=3DArial size=3D2>Okay, I got clarify to run on just =
plain logit=20
using Amelia. But is there some way to use amelia data sets with=20
relogit?</FONT></DIV>
<DIV><FONT face=3DArial size=3D2><BR>Traci</FONT></DIV></BODY></HTML>
------=_NextPart_000_0020_01C31238.0139A750--
For those of you who haven't used R before, I uploaded a brief
introduction to R which is written by Dan Ho for another class. It's under
``Handout'' on the course website.
A reminder that today I will hold an extra office hour right after the
lecture in my office at CBRSS. Tomorrow, I will have a regular office hour
from 4 to 6pm.
Kosuke
Here's our abstract. Any comments are greatly appreciated.
Best,
Dan and Colin
Crowded Cities, Crowded Streets: How Population Density Affects Local
Participation
Dan Hopkins and Colin Moore
Drawing on a survey of 29,000 Americans from communities throughout the United
States, this paper expands on J. Eric Oliver's (2000) argument that living in a
larger city decreases the probability that one will engage in a variety of
forms of local political participation. By shifting the causal emphasis from a
city¡¯s size to its population density, we illustrate that the effects of an
individual¡¯s environment depend crucially on the kind of participation in
question. As population density increases, an individual is 2.7% less likely
to attend a public meeting, but 2.0% more likely to participate in a
demonstration, boycott, or protest.
Hi,
I'm attempting to compute the robust standard errors for a weighted linear
regression. HCCM() only takes an un-weighted regression model. So, I'm
trying to compute the robust standard errors "directly". I'm comparing my
results to STATA, which computes the HC1 version of the HCCM automatically
for weighted linear regression. The following is my equation:
HC1 = N/(N-K)*inv(X'X)*(X'diag[e_i^2]X)*inv(X'X)
However, I'm not sure how to include the "weighted" part into the
calculation as there is nothing in the above equation that holds the
information about the weights of each observation.
I tried multiplying X by w, the vector of weights. This takes the standard
errors closer to what STATA produces, but I'm not sure if this is valid.
So, how would I incorporate the 'weighting' part of regression into the
heteroscedastic consistent covariance matrix?
Thanks,
John.
Here is the instruction about how to install R packages on ice matchines.
Kosuke
1. Create a library directory in your home directory by the following
commands.
ice% cd
ice% mkdir .R
ice% mkdir .R/library
2. Add the following line to .cshrc file. This defines the local library
directory you just created.
setenv R_LIBS ~/.R/library
3. Create links to all library files that already exist on ice servers by
the following command.
ice% ln -s /usr/lib/R/library/* .R/library/
4. Logout from your account or type the following command to make this
change effective.
ice% source .cshrc
Now, you can install additional R packages in your ice account. For
example, if you want to install the package called "car" at CRAN. Do the
following.
1. Start R.
2. install.packages("car")
If the pakcages are not at CRAN, you need to download it to your local
direcotry and then type at the command prompt
R INSTALL packagename