Statistical Methods for Detecting Anomalous Voting Patterns: A Case Study

Abstract

Both observational and laboratory studies suggest that human beings find it difficult to fabricate truly random numbers. Any sufficiently large set of nominally random number sequences, irrespective of its source, may be analyzed to determine if the criterion of randomness is satisfied. Statistical methods have, therefore, been applied forensically to detect anomalies in accounting data, scientific data sets, and voting data. As a case study, we apply methods developed by Beber and Scacco to analyze polling station counts in Helmand province for the four leading candidates in the 2009 Afghanistan presidential election. In this election, there were allegations of extensive electoral fraud, and nineteen percent of all votes cast were ultimately discarded. The pattern of statistical anomalies present in the raw polling station data is consistent with the findings of the post-election audit conducted by the Afghanistan Electoral Complaints Commission. The strengths and weaknesses of statistical methods for anomaly detection are discussed, as are possible motives for perpetuating voter fraud in elections that pose little uncertainty in projected outcomes.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 23, 2011
Accession Number
ADA550309

Entities

People

  • Roger Hillson

Organizations

  • United States Naval Research Laboratory

Tags

Communities of Interest

  • Biomedical

DTIC Thesaurus Topics

  • Afghanistan
  • Anomaly Detection
  • Case Studies
  • Change Detection
  • Data Analysis
  • Data Sets
  • Demographic Cohorts
  • Detection
  • Elections
  • Explosive Devices
  • Feedback
  • Frequency
  • Goodness Of Fit Tests
  • Improvised Explosive Devices
  • Information Science
  • Statistical Analysis
  • Statistical Tests

Readers

  • Educational Psychology
  • Political Violence and Terrorism Studies.
  • Regression Analysis.