Dear Ben,
Thanks for using Clarify. You are indeed correct; the code we give at the website is incomplete, and would have to be modified to get figure 1 from the paper. The actual code is available in our replication data set, available at http://gking.harvard.edu/data.shtml. For your convenience, I have included the Stata code that you are looking for below. As you will see, this code, in contrast with that on the web, produces 99% CI's. Sorry for the confusion.
Good luck with your research!
Best,
Jason Wittenberg
version 6.0
set mem 10m /* allocate RAM */
use turnout, clear /* load dataset */
estsimp logit turnout educate age agesqrd income white /* run logit */
setx (income white) mean /* hold inc,wt at means */
set more off /* let output scroll */
generate plo12 = . /* lower bound if ed=12 */
generate phi12 = . /* upper bound if ed=12 */
generate plo16 = . /* lower bound if ed=16 */
generate phi16 = . /* upper bound if ed=16 */
generate ageaxis = _n+17 in 1/78 /* contains 18,19,..,95 */
setx mean /* set all vars to mean */
local a = 18 /* begin with 18 ys old */
while `a' <= 95 { /* go to 95 years old */
display "Simulating for age = `a'" /* monitor our progress */
setx educate 12 age `a' agesqrd `a'^2*.01 /* set age, educ=12 ys */
simqi, prval(1) genpr(edis12) /* save Pr(Y=1|educ=12) */
_pctile edis12, p(.5,99.5) /* fetch percentiles */
replace plo12 = r(r1) if ageaxis==`a' /* save lo bound of ci */
replace phi12 = r(r2) if ageaxis==`a' /* save hi bound of ci */
setx educate 16 /* change educ to 16ys */
simqi, prval(1) genpr(edis16) /* save Pr(Y=1|educ=16) */
_pctile edis16, p(.5,99.5) /* fetch percentiles */
replace plo16 = r(r1) if ageaxis==`a' /* save lo bound of ci */
replace phi16 = r(r2) if ageaxis==`a' /* save hi bound of ci */
drop edis12 edis16 /* clean-up our mess */
local a = `a' + 1 /* advance age 1 year */
}
sort ageaxis
set textsize 130
graph plo12 phi12 plo16 phi16 ageaxis, s(iiii) c(||||) /*
*/ xlabel(18,24,30,36,42,48,54,60,66,72,78,84,90,95) /*
*/ ylabel(.2,.4,.6,.8,1) l1(Probability of Voting) /*
*/ gap(3) b2(Age of Respondent)
-----Original Message-----
From: Gary King [mailto:king@harvard.edu]
Sent: Wed 1/22/2003 6:33 PM
To: Ben Bishin
Cc: Mike Tomz; Jason Wittenberg
Subject: Re: A Clarify question
Thanks Ben. I'm CCing my coauthors who know more about the stata loops.
Gary
: Gary King, King(a)Harvard.Edu http://GKing.Harvard.Edu :
: Center for Basic Research Direct (617) 495-2027 :
: in the Social Sciences Assistant (617) 495-9271 :
: 34 Kirkland Street, Rm. 2 HU-MIT DC (617) 495-4734 :
: Harvard U, Cambridge, MA 02138 eFax (928) 832-7022 :
On Wed, 22 Jan 2003, Ben Bishin wrote:
> Hi Gary,
>
>
> I hope all is well in Boston.
>
> I write with a Clarify question. Well actually its about the documentation
> for the Figure 1 in the "Software for Interpreting and Presenting
> Statistical Results" handbook that goes with the software. Im probably
> missing something as I haven't graphed before in Stata, but looking at the
> code on page 21 (I also found the same discussion
> here: http://gking.harvard.edu/clarify/docs/node24.html ) it doesnt seem
> to provide the CI's or estimates for both groups depicted in the figure in
> the AJPS paper. From my reading, it only generates the predicted values
> and CI's for the High School group (the value for the education variable
> seems fixed at 12). It seems like there should be a loop for the College
> group too. Am I missing something? If not is it possible to get the actual
> code from the paper?
>
> Also, this is lint but it appears that this code produces 95% CIs while the
> paper reports 99%CIs (Figure 1 on page 355).
>
> One last question: Im working on this for a sociology paper Im writing
> with a colleague, and was wondering if you are aware of any papers using
> clarify in the sociology literature?
>
> Thanks!
>
> Ben
>
> PS: I love this program.
>
>
>
> Benjamin G. Bishin
>
> Assistant Professor
> Department of Political Science
> University of Miami
> 314 Jenkins Building
> Coral Gables, FL 33124-6534
>
> Phone: (305) 284-1737
> Fax: (305) 284-3636
> http://homer.bus.miami.edu/~bbishin/
>
--
Clarify mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.hmdc.harvard.edu/?info=clarify
Clarify is now Amelia-ready in the sense that there are facilities within
it to use multiply imputed data sets. So you don't need other software
any longer. For details of how to use this, see
http://gking.harvard.edu/clarify/docs/node26.html
Thanks for your interest.
Gary King
: Gary King, King(a)Harvard.Edu http://GKing.Harvard.Edu :
: Center for Basic Research Direct (617) 495-2027 :
: in the Social Sciences Assistant (617) 495-9271 :
: 34 Kirkland Street, Rm. 2 HU-MIT DC (617) 495-4734 :
: Harvard U, Cambridge, MA 02138 eFax (928) 832-7022 :
On Mon, 2 Dec 2002 TxLiberal(a)aol.com wrote:
> I am in a Mulitvariate class at the University of Houston. We are using
> Stata, Clarigy and Amelia and were told to use the MIEST command in
> Stata. However, we cannot find where to download the MIEST software.
> Any assistance you could offer would be greatly appreciated.
>
> Thanks,
>
> Tom
>
--
Clarify mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.hmdc.harvard.edu/?info=clarify
Yes, that's the plan, but probably not immediately. (There will be a
program available sometime soon in R that does all of what Clarify and
Relogit does and more.)
But I think you ought to be able to use the relogit program to produce
simulations and any quantity of interest you might like from Clarify.
Gary
On Fri, 29 Nov 2002, Linne, Thomas wrote:
>
> Dear Professor King,
>
> I'm using Clarify 2.0 and ReLogit 1.1 for Stata.
> Especially Clarify is very handy when it comes to visualising the
> results of
> a probit or logit model.
>
> Unfortunately, the estsimp command in current version of Clarify
> doesn't support the relogit command.
>
> For the presentation of the results of a rare event logit model I've
> therefore
> used the logit command. Since there are differences in the estimated
> coefficients
> I doubt that this workaround will suffice for a journal publication.
>
> I'm wondering if there are any plans to extend the estsimp command in
> order to
> make it workable with the relogit comand?
>
> Regards
>
> -Thomas.
>
>
> ______________________________________________________________________
>
> Thomas Linne
> Institute for Economic Research
> Dept. Central and Eastern Europe
> P.O. Box 11 03 61
> 06017 Halle/Saale
> Germany
>
> email: Thomas.Linne(a)iwh-halle.de Tel.: 00 49 - (0)345 - 77 53 834
> Web: www.iwh-halle.de Fax: 00 49 - (0)345 - 77 53 820
>
--
Clarify mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.hmdc.harvard.edu/?info=clarify
if you have missing values on some but not all of the items in the scale,
then impute first and sum at the end of the process of imputation.
Gary King
: Gary King, King(a)Harvard.Edu http://GKing.Harvard.Edu :
: Center for Basic Research Direct (617) 495-2027 :
: in the Social Sciences Assistant (617) 495-9271 :
: 34 Kirkland Street, Rm. 2 HU-MIT DC (617) 495-4734 :
: Harvard U, Cambridge, MA 02138 eFax (928) 832-7022 :
On Fri, 22 Nov 2002, Michael Sauer wrote:
> I have downloaded both Amelia for Windows and Clarify for Stata and I
> have a questions about something I am not clear about.
>
> I have survey data that I had intended to create summated scales on
> based on scales that had been defined by the survey author. Before I
> can create these scales, it seems to me that I need to impute the
> missing values for the individual survey items and run my ordered
> probit using the scale calculated above as the dependent variable.
>
> At what point in the Amelia (imputing the missing observations) to
> Clarify process would I create my summated scales?
>
> I would really appreciate any help you can provide.
>
> Michael L. Sauer
> Associate Vice President for Administration
> University of Nevada Las Vegas
> 4505 Maryland Parkway
> Las Vegas, NV 89154
> Phone: (702) 895-1073
> Fax: (702) 895-4929
> email: msauer(a)unlv.edu
>
--
Clarify mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.hmdc.harvard.edu/?info=clarify
Hi Gary--
I recently started using Amelia (Windows version) to generate
multiply-imputed datasets and Clarify package / -miest- to run statistical
models and pull the parameter estimates together. I am still learning how to
use it.
I have a couple of question:
1) Is it possible to set a seed number, both in Amelia and in Clarify, so
results can be exactly reproduced?
2) Is it possible to have Amelia save the settings of any run in a file, to
document these settings?
3) (Just a wish list item. I don't think this is possible). Is it possible
to run Amelia from a batch file, something similar to Stata's -do- files? It
would save so much typing...
4) (Another item to add to the wish list). Do you have any plans to have an
Amelia version for Stata? Or is it just unpractical? (I would then have
access to all the parts of Amelia which currently only work with the Gauss
version, such as the use of other MI algorithms, and the use of graphs.)
Thank you for your help,
--Estie Hudes.
Estie Sid Hudes, PhD MPH
University of California
Prevention Sciences Group &
Center for AIDS Prevention Studies
74 New Montgomery St., Suite 600, Box 0886
S.F. CA 94105
phone: 415.597.9126
fax: 415.597.9194
email: ehudes(a)psg.ucsf.edu
Web: http://www.epibiostat.ucsf.edu/capsweb/people/hudesbio.html
--
Clarify mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.hmdc.harvard.edu/?info=clarify
Frustrated a little by the difficulty of running -clarify- with multiple
variables and many multiples of setx, particularly involving Stata's xi
prefix, I have written a wrapper for the clarify series that automates this
process.
The program and help files are available from Stata (thanks to Kit Baum) by
typing -ssc install qsim-
Title: QSIM
qsim provides a wrapper for the clarify series of program that automates
simulation , particularly with dummy variables produced by Stata's xi prefix.
Without qsim each of the 3 clarify programs has to be run separately for
each level of the xi: categorical variable. qsim can be placed into one or
more foreach loops and a large series of
simulations can be run at one time.
The output can be pasted into a text editor. Marking E( or Pr( identifies
the line with the simulation data output. This line then can be cut to
Excel (or similar programs) to make an output table. The whole process
takes less than a minute.
Fred Wolfe
----------------------------------------------------------------------------
------------------------
Fred Wolfe Tel
(316) 263-2125
National Data Bank for Rheumatic Diseases Fax (316) 263-0761
Wichita,
Kansas fwolfe(a)arthritis-research.org
----------------------------------------------------------------------------
-------------------------
clarify mailing list served by Harvard-MIT Data Center
List Address: clarify(a)latte.harvard.edu
Subscribe/Unsubscribe: http://lists.hmdc.harvard.edu/listis.cgi?info=clarify
Hi Vincent,
Thanks for using Clarify. Let me answer your questions in the order in which
you ask them.
1. Unfortunately SETX does not permit you to specify different values for
different equations in a multi-equation model. We are aware of this
drawback, and will try to address it in our next update.
2. There are two possible outcomes when you type SETX without anything else
on the line. If you have not typed SETX before, you will receive a message
saying, "No values have been set for the explanatory variables." This does
not mean that the values have been set to zero, but that you have not yet
set any values for the explanatory variables. If you type SETX without
anything else *after* having set the values, then you will see the values
that you have set the X's to on the screen. To set a variable called X1 to
0, type "setx x1 0"
3. To set all variables to their means, type "setx mean". To set only
variable X1 to its mean, type "setx x1 mean"
4. I am not sure what you are after in question 4. Perhaps you should
elaborate a little. Clarify does not include any graphics capability. The
replication information for our paper, including any graphics code, can be
downloaded from http://gking.harvard.edu.
Good luck with your research!
Jason Wittenberg
-----Original Message-----
From: vincent
To: clarify(a)latte.harvard.edu
Cc: yc258(a)columbia.edu
Sent: 8/31/02 11:14 PM
Subject: [clarify] syntax of using SURE in clarify
Dear Prof King, Prof Tomz and Prof Wittenberg:
> thank you for posting the software on the web. it is very useful!
>
> i am using your "clarify" for seemingly
> unrelated regression. can you kindly let me know the following problem
on
> syntax?
>
> since i have some explantory variables are common in those equations,
i
> have the following questions of analysis specifications:
>
> 1. how can i specify different value of those same variables
> simultanously appear in different equations when using "setx" ?
>
> 2. is it meant that i am making all the explantory variables set to
> zero when i typed "setx" only.
>
> 3. is it meant that all other variables are set to their "mean" while
i
> just specified only one variable?
>
> 4. how can i perform the first differencing and the graph of the
seeming
> uncorrelated regression.
yours
yiu por
clarify mailing list served by Harvard-MIT Data Center
List Address: clarify(a)latte.harvard.edu
Subscribe/Unsubscribe: http://lists.hmdc.harvard.edu/listis.cgi?info=clarify
i've been running some first differences in clarify
that are based on multinomial logit models where the
dependent variable has a base category 0 and
comparison categories -1 and 1. i've noticed that
the first difference for the highest category (in this
case, 1) always has a *tiny* standard error,
regardless of the error in the corresponding
coefficient. i can reverse the order of the variable
(1 becomes -1 and vice versa) and the highest category
still gets the smallest s.e. i can obviously code my
dep var to give the highest value to the comparison
category so the results i report are more in line with
the coefficients, but the phenomenon has concerned me.
does anyone know why this happens? does it have to
do with the fact that the results of any given
category are always perfectly predicted by the results
of the other two? or is it something about my data?
confused,
eric mcghee
__________________________________________________
Do You Yahoo!?
Yahoo! Finance - Get real-time stock quotes
http://finance.yahoo.com
clarify mailing list served by Harvard-MIT Data Center
List Address: clarify(a)latte.harvard.edu
Subscribe/Unsubscribe: http://lists.hmdc.harvard.edu/listis.cgi?info=clarify
