This zipped directory contains the raw data and the R programs and output files used to obtain the results reported in the Tables and Figures in the report on Franklin County precincts produced for the Democratic National Committee by Walter Mebane (franklin.pdf, June 18, 2005, ``Ohio 2004 Election: New Registrants, Provisional Ballots, Voting Machines, Turnout and Polls Open Elapsed Times in Franklin County Precincts''). That report is a supplement to the Ohio precincts study report produced for the Democratic National Committee by Walter Mebane and Michael Herron (Ohio_precincts.09june2005.pdf, ``Ohio 2004 Election: Turnout, Residual Votes and Votes in Precincts and Wards'').

The output files were produced by running R in BATCH mode in commands like

  R CMD BATCH --no-save franklin1.R

Hence the output file corresponding to program foo.R is foo.Rout.

The R program files that perform robust estimation of the overdispersed binomial regression model use a customized version of the package MultinomRob. That version of MultinomRob should be available soon through CRAN as Version 1.3 of the MultinomRob package.

Mebane, Walter R., Jr., and Jasjeet Sekhon. 2004b. "Multinomial Robust Regression (MultinomRob)." Package for R available from the Comprehensive R Archive Network (CRAN, http://cran.r-project.org/).

To execute the R programs that use MultinomRob, install the package and in the R programs change

        library(multinomRob, lib.loc="/home/wrm1/lib/R/");

to

        library(multinomRob);

Or change the line to point to the directory where the package is installed. If you rerun the programs some of the robust estimation results may differ slightly from the ones in the files included in this distribution. This is primarily because the robust estimation algorithm is stochastic (different runs of the GENOUD search algorithm used to find the LQD estimates of the dispersion may produce slightly different results).

Files included:

  -rw-rw----   1 wrm1 research     3342 Jun 20 16:13 README3
  -rw-rw----   1 wrm1 research     3989 Jun 18 00:43 franklin1.Rout
  -rw-rw----   1 wrm1 research    22462 Jun 18 00:43 franklinA.ps
  -rw-rw----   1 wrm1 research    22650 Jun 18 00:43 franklinB.ps
  -rw-rw----   1 wrm1 research    22438 Jun 18 00:43 franklinC.ps
  -rw-rw----   1 wrm1 research    62925 Jun 18 00:43 franklinD.ps
  -rw-rw----   1 wrm1 research     2936 Jun 18 00:42 franklin1.R
  -rw-rw----   1 wrm1 research     1367 Jun 17 17:41 franklin4.calc.Rout
  -rw-rw----   1 wrm1 research      477 Jun 17 17:41 franklin4.calc.R
  -rw-rw----   1 wrm1 research     9138 Jun 17 17:03 franklin4.Rout
  -rw-rw----   1 wrm1 research   736829 Jun 17 17:03 .Rdata.franklin4
  -rw-rw----   1 wrm1 research     2658 Jun 17 16:37 franklin4.R
  -rw-rw----   1 wrm1 research    15545 Jun 17 04:09 franklin3.Rout
  -rw-rw----   1 wrm1 research   865390 Jun 17 04:09 .Rdata.franklin3
  -rw-rw----   1 wrm1 research     3338 Jun 17 03:08 franklin3.R
  -rw-rw----   1 wrm1 research    29105 Jun 17 03:04 franklin2A.ps
  -rw-rw----   1 wrm1 research    22589 Jun 17 03:04 franklin2B.ps
  -rw-rw----   1 wrm1 research     6692 Jun 17 03:04 franklin2.Rout
  -rw-rw----   1 wrm1 research     2656 Jun 17 03:03 franklin2.R
  -rw-r--r--   1 wrm1 research    70849 Jun 16 17:43 franklinMchWrkProPollClose.csv

Walter Mebane
20 June 2005