\batchmode

















%

\documentclass[12pt]{article}
\usepackage{fullpage}
\usepackage{bib}
\usepackage{setspace}
\usepackage{endold}
\parskip=12pt

\makeatother
\newenvironment{tex2html_wrap}{}{}
\newwrite\lthtmlwrite
\def\lthtmltypeout#1{{\let\protect\string\immediate\write\lthtmlwrite{#1}}}%
\newbox\sizebox
\begin{document}
\pagestyle{empty}
{\newpage
\clearpage
\samepage \begin{centering}\vspace{.1in}
\begin{large}
Constituents and Knowledge of Senate Roll Call Votes
\end{large}
\vspace{.5in}

R. Michael Alvarez\\ 
{\em California Institute of Technology}\\ 
\emph{rma@hss.caltech.edu}\\ 
\vspace{.125in}
Paul Gronke\\ 
{\em Duke University}\\ 
\emph{gronke@acpub.duke.edu}\\ 
\vspace{.5in}

\today

\end{centering}
}

\setcounter{page}{1}
\stepcounter{paragraph}
\stepcounter{section}
\stepcounter{section}
\stepcounter{section}
\stepcounter{section}
\stepcounter{section}
{\newpage
\clearpage
\samepage \begin{table}\begin{center}

\label{t3-recall}
\begin{tabular}{lcc|cc}
Independent & \multicolumn{2}{c}{Gulf War Vote} & \multicolumn{2}{c}{Thomas 
Vote} \\ 
Variables   & Estimates & Effects               & Estimates & Effects \\  \hline 
Constant    & 20.4*     &                       & 42.1*     & \\ 
            & 9.8       &                       & 11.8      & \\ 
Media Attn. & -.002     & -.0006                & .18       & .07  \\ 
            & .16       & .06                   & .19       & .07  \\ 
Pol. Info   & .43*      & .16*                  & .09       & .04  \\ 
            & .19       & .07                   & .21       & .08  \\ 
Partisan    & .08       & .03                   & .07       & .03  \\ 
Agreement   & .08       & .03                   & .10       & .04  \\ 
Extremity   & -.008*    & -.003*                & -.01*     & -.004*  \\ 
            & .003      & .001                  & .003      & .001  \\ 
Tenure      & -.01*     & -.004*                & -.02*     & -.008*  \\ 
            & .005      & .002                  & .006      & .002  \\ 
Reelection  & .02       & .009                  & -.24*     & -.10*  \\ 
            & .10       & .04                   & .12       & .05  \\ 
$\chi^{2}$  & 19.12     &                       & 30.4      &  \\ 
N           &  958      &                       & 722       &  \\  \hline
\end{tabular}
\parbox{5in}{Note:  For each model, we provide maximum-likelihood probit
estimates in the left-most column, with their standard errors below.  The
next column for each model presents the marginal effects of each 
variable; the marginal effects for the continuous variables are the difference
in probability achieved by moving the variable one unit from the sample 
mean, while for the binary variables it is the difference in probability
achieved by moving the variable from zero to one.  * denotes and 
estimate statistically significant at the p=.05 level, one-tailed test.}
\end{center}
\end{table}
}

{\newpage
\clearpage
\samepage \begin{table}\begin{center}

\label{t3-guess}
\begin{tabular}{lcc|cc}
Independent & \multicolumn{2}{c}{Gulf War Vote} & \multicolumn{2}{c}{Thomas 
Vote} \\ 
Variables   & Estimates & Effects               & Estimates & Effects \\  \hline 
Constant    & -12.1*    &                       & -3.8      &  \\ 
            & 7.1       &                       & 7.1       &  \\ 
Media Attn. & -.03      & -.01                  & .06       & .02  \\ 
            & .09       & .04                   & .10       & .04  \\ 
Pol. Info   & -.13      & -.05                  & .09       & .03  \\ 
            & .12       & .05                   & .13       & .05  \\ 
Partisan    & .14*      & .06*                  & .12*      & .05*  \\ 
Agreement   & .06       & .02                   & .06       & .02  \\ 
Extremity   & .0005     & .0002                 & -.002     & -.0009  \\ 
            & .002      & .0007                 & .002      & .0007  \\ 
Tenure      & .006*     & .002*                 & .002      & .0008  \\ 
            & .004      & .001                  & .004      & .001  \\ 
Reelection  & .01       & .005                  & -.10      & -.04  \\ 
            & .07       & .03                   & .07       & .03  \\ 
$\chi^{2}$  &  11.06    &                       & 9.11      &  \\ 
N           &  2165     &                       & 2073      &  \\  \hline
\end{tabular}
\parbox{5in}{Note:  For each model, we provide maximum-likelihood probit
estimates in the left-most column, with their standard errors below.  The
next column for each model presents the marginal effects of each 
variable; the marginal effects for the continuous variables are the difference
in probability achieved by moving the variable one unit from the sample 
mean, while for the binary variables it is the difference in probability
achieved by moving the variable from zero to one.  * denotes and 
estimate statistically significant at the p=.05 level, one-tailed test.}
\end{center}
\end{table}
}

\stepcounter{section}

\end{document}