PERFECT Project: Unbiased Visual Auditing

One of the goals we are pursuing in our research is the development of semi-automated tools to help eliminate human bias in auditing election results. Bias arises because people naturally have an inherent preference for particular candidates or issues.

We are planning to conduct two types of experiments that may lead to simple ways of avoiding such bias. Both experiments will make use of the ballot images generated using other software tools we have developed.

The first set of experiments will consist of comparing human vote counts using context versus human counts without using context (oddly called ``blind recording''). Context is present when enough of the ballot is visible to the human counter to identify the party, candidate, or proposition for which a vote was cast. Context is absent when only the mark area is displayed. The screen snapshots below illustrate the steps required to eliminate context and enable blind audits.

Unbiased auditing 1
(a) Original ballot image.

Unbiased auditing 2
(b) Candidate labels obscured.

Unbiased auditing 3
(c) Voter’s selections randomly shuffled within races.

PERFECT is an acronym that stands for "Paper and Electronic Records for Elections: Cultivating Trust." PERFECT is a multidisciplinary research effort aimed at studying the reliable processing of paper ballots and other hardcopy election records. Participating institutions include Lehigh University, Boise State University, Muhlenberg College, and Rensselaer Polytechnic Institute. Click here to return to the PERFECT homepage.

NSF logo
PERFECT is funded in part by the National Science Foundation under award numbers NSF-0716368, NSF-0716393, NSF-0716647, NSF-0716543. Any opinions, findings, and conclusions or recommendations expressed on this website are the investigators' and do not necessarily reflect those of the National Science Foundation.