Walter Mebane - Academia.edu (original) (raw)

Papers by Walter Mebane

DOAJ (DOAJ: Directory of Open Access Journals), Aug 1, 2011

SSRN Electronic Journal, 2017

Computational Communication Research

source files (R and .dta), output (.Rout and .RData), graphs (pdf)

Depends R (> = 1.7.1), rgenoud (> = 2.9), MASS (> = 7.1-8), mvtnorm (> = 0.6-3)

Proceedings of the …, 2001

... Wand, Jonathan NA, Kenneth Shotts, Jasjeet S. Sekhon, Walter R. Mebane, Jr., Michael Herron, ... more ... Wand, Jonathan NA, Kenneth Shotts, Jasjeet S. Sekhon, Walter R. Mebane, Jr., Michael Herron, and Henry E. Brady. 2001. The Butterfly Did It: The Aberrant Vote for Buchanan in Palm Beach County, Florida. American Political Science Review. Forth-coming. ...

There is an acute need for methods of detecting and investigating fraud in elections, because the... more There is an acute need for methods of detecting and investigating fraud in elections, because the consequences of electoral fraud are grave for democratic stability and quality. When the electoral process is compromised by fraud, intimidation, or even violence, elections can become corrosive and destabilizing—sapping support for democratic institutions; inflaming suspicion; and stimulating demand for extra-constitutional means of pursuing political agendas, including violence. Accurate information about irregularities can help separate false accusations from evidence of electoral malfeasance. Accurate information about the scope of irregularities can also provide a better gauge of election quality. Finally, accurate information about the geographic location of malfeasance—the locations where irregularities occurred and how they cluster—can allow election monitors and pro-democracy organizations to focus attention and resources more efficiently and to substantiate their assessments o...

The "EFToolkit" package helps to implement election forensics analysis using different ... more The "EFToolkit" package helps to implement election forensics analysis using different methods and estimators. It is partly built on the code from "Election Forensics Toolkit" web application sponsored by the USAID and developed by Walter Mebane and Kirill Kalinin.

computer codes (programming language R), batch run output files and data that produced informatio... more computer codes (programming language R), batch run output files and data that produced information in all tables and figures in the paper and supplementary information

Supervised machine learning methods are increasingly employed in political science. Such models r... more Supervised machine learning methods are increasingly employed in political science. Such models require costly manual labeling of documents. In this paper we introduce active learning, a framework in which data to be labeled by human coders are not chosen at random but rather targeted in such a way that the required amount of data to train a machine learning model can be minimized. We study the benefits of active learning using text data examples. We perform simulation studies that illustrate conditions where active learning can reduce the cost of labeling text data. We perform these simulations on three corpora that vary in size, document length and domain. We find that in cases where the document class of interest is not balanced, researchers can label a fraction of the documents one would need using random sampling (or `passive' learning) to achieve equally performing classifiers. We further investigate how varying levels of inter-coder reliability affect the active learning ...

this paper we do not report estimates for the covariance structure model. Instead we present prel... more this paper we do not report estimates for the covariance structure model. Instead we present preliminary results for the United States during years 1984--88 using a generalized linear model and quasi-likelihood estimation with American National Election Studies (ANES) data, U.S. local area economic data and PANDA project events data. The preliminary results suggest that the major features of the covariance structure model ought to apply well to protest events. They also give strong support for the central predictions of our low-dimension, sunspot theory

Do American Midterm House Electors Coordinate? Electors (eligible voters) have been coordinating ... more Do American Midterm House Electors Coordinate? Electors (eligible voters) have been coordinating their turnout decisions and vote choices for the House of Representatives in recent midterm elections. Coordination is a noncooperative rational expectations equilibrium among electors. The theory allows some electors not to care about policy and admits House districts in which a candidate is running unopposed. Stochastic choice models estimated using individual-level data from NES surveys of years 1978--1998 support both coordination and a model in which electors act nonstrategically. Parameter estimates for the coordinating model describe policy-moderating behavior, but estimates for the nonstrategic model mostly do not. Both models show that many voters had policy-related incentives to change their votes between presidential year and midterm---enough to produce a midterm cycle in which the President's party usually loses vote share at midterm. Midterm electors have expected the ...

DOAJ (DOAJ: Directory of Open Access Journals), Aug 1, 2011

SSRN Electronic Journal, 2017

Computational Communication Research

source files (R and .dta), output (.Rout and .RData), graphs (pdf)

Depends R (> = 1.7.1), rgenoud (> = 2.9), MASS (> = 7.1-8), mvtnorm (> = 0.6-3)

Proceedings of the …, 2001

... Wand, Jonathan NA, Kenneth Shotts, Jasjeet S. Sekhon, Walter R. Mebane, Jr., Michael Herron, ... more ... Wand, Jonathan NA, Kenneth Shotts, Jasjeet S. Sekhon, Walter R. Mebane, Jr., Michael Herron, and Henry E. Brady. 2001. The Butterfly Did It: The Aberrant Vote for Buchanan in Palm Beach County, Florida. American Political Science Review. Forth-coming. ...

There is an acute need for methods of detecting and investigating fraud in elections, because the... more There is an acute need for methods of detecting and investigating fraud in elections, because the consequences of electoral fraud are grave for democratic stability and quality. When the electoral process is compromised by fraud, intimidation, or even violence, elections can become corrosive and destabilizing—sapping support for democratic institutions; inflaming suspicion; and stimulating demand for extra-constitutional means of pursuing political agendas, including violence. Accurate information about irregularities can help separate false accusations from evidence of electoral malfeasance. Accurate information about the scope of irregularities can also provide a better gauge of election quality. Finally, accurate information about the geographic location of malfeasance—the locations where irregularities occurred and how they cluster—can allow election monitors and pro-democracy organizations to focus attention and resources more efficiently and to substantiate their assessments o...

The "EFToolkit" package helps to implement election forensics analysis using different ... more The "EFToolkit" package helps to implement election forensics analysis using different methods and estimators. It is partly built on the code from "Election Forensics Toolkit" web application sponsored by the USAID and developed by Walter Mebane and Kirill Kalinin.

computer codes (programming language R), batch run output files and data that produced informatio... more computer codes (programming language R), batch run output files and data that produced information in all tables and figures in the paper and supplementary information

Supervised machine learning methods are increasingly employed in political science. Such models r... more Supervised machine learning methods are increasingly employed in political science. Such models require costly manual labeling of documents. In this paper we introduce active learning, a framework in which data to be labeled by human coders are not chosen at random but rather targeted in such a way that the required amount of data to train a machine learning model can be minimized. We study the benefits of active learning using text data examples. We perform simulation studies that illustrate conditions where active learning can reduce the cost of labeling text data. We perform these simulations on three corpora that vary in size, document length and domain. We find that in cases where the document class of interest is not balanced, researchers can label a fraction of the documents one would need using random sampling (or `passive' learning) to achieve equally performing classifiers. We further investigate how varying levels of inter-coder reliability affect the active learning ...

this paper we do not report estimates for the covariance structure model. Instead we present prel... more this paper we do not report estimates for the covariance structure model. Instead we present preliminary results for the United States during years 1984--88 using a generalized linear model and quasi-likelihood estimation with American National Election Studies (ANES) data, U.S. local area economic data and PANDA project events data. The preliminary results suggest that the major features of the covariance structure model ought to apply well to protest events. They also give strong support for the central predictions of our low-dimension, sunspot theory

Do American Midterm House Electors Coordinate? Electors (eligible voters) have been coordinating ... more Do American Midterm House Electors Coordinate? Electors (eligible voters) have been coordinating their turnout decisions and vote choices for the House of Representatives in recent midterm elections. Coordination is a noncooperative rational expectations equilibrium among electors. The theory allows some electors not to care about policy and admits House districts in which a candidate is running unopposed. Stochastic choice models estimated using individual-level data from NES surveys of years 1978--1998 support both coordination and a model in which electors act nonstrategically. Parameter estimates for the coordinating model describe policy-moderating behavior, but estimates for the nonstrategic model mostly do not. Both models show that many voters had policy-related incentives to change their votes between presidential year and midterm---enough to produce a midterm cycle in which the President's party usually loses vote share at midterm. Midterm electors have expected the ...