Security analysis of the diebold AccuVote-TS voting machine
EVT'07 Proceedings of the USENIX Workshop on Accurate Electronic Voting Technology
Three voting protocols: ThreeBallot, VAV, and twin
EVT'07 Proceedings of the USENIX Workshop on Accurate Electronic Voting Technology
On auditing elections when precincts have different sizes
EVT'08 Proceedings of the conference on Electronic voting technology
EVT'08 Proceedings of the conference on Electronic voting technology
Some consequences of paper fingerprinting for elections
EVT/WOTE'09 Proceedings of the 2009 conference on Electronic voting technology/workshop on trustworthy elections
Weight, weight, don't tell me: using scales to select ballots for auditing
EVT/WOTE'09 Proceedings of the 2009 conference on Electronic voting technology/workshop on trustworthy elections
On the security of election audits with low entropy randomness
EVT/WOTE'09 Proceedings of the 2009 conference on Electronic voting technology/workshop on trustworthy elections
Bubble trouble: off-line de-anonymization of bubble forms
SEC'11 Proceedings of the 20th USENIX conference on Security
| Hi-index | 0.00 |
We consider the problem of statistical sampling for auditing elections, and we develop a remarkably simple and easily-calculated upper bound for the sample size necessary for determining with probability at least c if a given set of n objects contains fewer than b "bad" objects. While the size of the optimal sample drawn without replacement can be determined with a computer program, our goal is to derive a highly accurate and simple formula that can be used by election officials equipped with only a hand-held calculator. We actually develop several formulae, but the one we recommend for use in practice is: U3(n, b, c) = ⌈(n - (b - 1)/2) ċ (1 - (1 - c)1/b)⌉ = ⌈(n - (b - 1)/2) ċ (1 - exp(ln(1 - c)/b))⌉ As a practical matter, this formula is essentially exact: we prove that it is never too small, and empirical testing for many representative values of n ≤ 10,000, and b ≤ n/2, and c ≤ 0.99 never finds it more than one too large. Theoretically, we show that for all n and b this formula never exceeds the optimal sample size by more than 3 for c ≤ 0.9975, and by more than (-ln(1 - c))/2 for general c.