Search for: All records

Abstract ONEAudit provides more efficient risk-limiting audits than other extant methods when the voting system cannot report a cast-vote record linked to each cast card. It obviates the need for re-scanning; it is simpler and more efficient than ‘hybrid’ audits; and it is far more efficient than batch-level comparison audits. There may be room to improve the efficiency of ONEAudit further by tuning the statistical tests it uses and by using stratified sampling. We show that tuning the tests by optimizing for the reported batch-level tallies or integrating over a distribution reduces expected workloads by 70–85% compared to the current ONEAudit implementation across a range of simulated elections. The improved tests reduce the expected workload to audit the 2024 Mayoral race in San Francisco, California, by half—from about 200 cards to about 100 cards. In contrast, stratified sampling does not help: it increases workloads by about 25% on average. 

One approach to risk-limiting audits (RLAs) compares ran- domly selected cast vote records (CVRs) to votes read by human auditors from the corresponding ballot cards. Historically, such methods reduce audit sample sizes by considering how each sampled CVR di!ers from the corresponding true vote, not merely whether they di!er. Here we investigate the latter approach, auditing by testing whether the total number of mismatches in the full set of CVRs exceeds the minimum number of CVR errors required for the reported outcome to be wrong (the “CVR margin”). This strategy makes it possible to audit more social choice functions and simplifies RLAs conceptually, which makes it easier to explain than some other RLA approaches. The cost is larger sample sizes. “Mismatch-based RLAs” only require a lower bound on the CVR margin, which for some social choice functions is easier to calculate than the e!ect of particular errors. When the population rate of mismatches is low and the lower bound on the CVR margin is close to the true CVR margin, the increase in sample size is small. However, the increase may be very large when errors include errors that, if corrected, would widen the CVR margin rather than narrow it; errors a!ect the margin between candidates other than the reported winner with the fewest votes and the reported loser with the most votes; or errors that a!ect di!erent margins. 

Abstract AWAIRE is one of two extant methods for conducting risk-limiting audits of instant-runoff voting (IRV) elections. In principle AWAIRE can audit IRV contests with any number of candidates, but the original implementation incurred memory and computation costs that grew superexponentially with the number of candidates. This paper improves the algorithmic implementation of AWAIRE in three ways that make it practical to audit IRV contests with 55 candidates, compared to the previous 6 candidates. First, rather than trying from the start to rule out all candidate elimination orders that produce a different winner, the algorithm starts by considering only the final round, testing statistically whether each candidate could have won that round. For those candidates who cannot be ruled out at that stage, it expands to consider earlier and earlier rounds until either it provides strong evidence that the reported winner really won or a full hand count is conducted, revealing who really won. Second, it tests a richer collection of conditions, some of which can rule out many elimination orders at once. Third, it exploits relationships among those conditions, allowing it to abandon testing those that are unlikely to help. We provide real-world examples with up to 36 candidates and synthetic examples with up to 55 candidates, showing how audit sample size depends on the margins and on the tuning parameters. An open-source Python implementation is publicly available. 

U.S. elections rely heavily on computers such as voter registration databases, electronic pollbooks, voting machines, scanners, tabulators, and results reporting websites. These introduce digital threats to election outcomes. Risk-limiting audits (RLAs) mitigate threats to some of these systems by manually inspecting random samples of ballot cards. RLAs have a large chance of correcting wrong outcomes (by conducting a full manual tabulation of a trustworthy record of the votes), but can save labor when reported outcomes are correct. This efficiency is eroded when sampling cannot be targeted to ballot cards that contain the contest(s) under audit. If the sample is drawn from all cast cards, then RLA sample sizes scale like the reciprocal of the fraction of ballot cards that contain the contest(s) under audit. That fraction shrinks as the number of cards per ballot grows (i.e., when elections contain more contests) and as the fraction of ballots that contain the contest decreases (i.e., when a smaller percentage of voters are eligible to vote in the contest). States that conduct RLAs of contests on multi-card ballots or RLAs of small contests can dramatically reduce sample sizes by using information about which ballot cards contain which contests—by keeping track of card-style data (CSD). For instance, CSD reduce the expected number of draws needed to audit a single countywide contest on a 4-card ballot by 75%. Similarly, CSD reduce the expected number of draws by 95% or more for an audit of two contests with the same margin on a 4-card ballot if one contest is on every ballot and the other is on 10% of ballots. In realistic examples, the savings can be several orders of magnitude. 

The “job talk” is a standard element of faculty recruiting. How audiences treat candidates for faculty positions during job talks could have disparate impact on protected groups, including women. We annotated 156 job talks from five engineering and science departments for 13 categories of questions and comments. All departments were ranked in the top 10 by US News & World Report. We find that differences in the number, nature, and total duration of audience questions and comments are neither material nor statistically significant. For instance, the median difference (by gender) in the duration of questioning ranges from zero to less than two minutes in the five departments. Moreover, in some departments, candidates who were interrupted more often were more likely to be offered a position, challenging the premise that interruptions are necessarily prejudicial. These results are specific to the departments and years covered by the data, but they are broadly consistent with previous research, which found differences comparable in magnitude. However, those studies concluded that the (small) differences were statistically significant. We present evidence that the nominal statistical significance is an artifact of using inappropriate hypothesis tests. We show that it is possible to calibrate those tests to obtain a properP-value using randomization.