Voting Machines, Ballots Should Be Designed, Tested Based on Human-Factors Principles, UB Engineer Says

Release Date: November 17, 2000 This content is archived.

Print

BUFFALO, N.Y. -- The same principles that ensure user-friendly designs in products ranging from refrigerators to computers to dashboards on automobiles should be applied to the design of both paper and machine election ballots, according to a UB professor of industrial engineering.

"No matter what the system is, you need to apply good design principles to it," said Ann Bisantz, Ph.D., assistant professor of industrial engineering at UB.

In the wake of the experience of Florida voters who complained about ballots used in some counties in this year's presidential election, she said, "some have charged that voters didn't pay sufficient attention while they were casting their ballots, that maybe it's all just human error."

Finding the root cause of human errors in a specific system is the focus of human-factors professionals like Bisantz.

"If human errors can happen in a complex system like an aircraft with a highly skilled individual like an airline pilot," she noted, "they certainly can happen in a voting booth with a person who votes once every four years or even less often."

Bisantz recently received a $300,000 grant from the National Science Foundation to study human decision-making in complex environments, such as manufacturing, aviation, transportation and the military.

Human-factors research focuses on helping workers make better decisions in high-risk sectors.

Application of the same principles to the voting process, including a detailed testing of the design, would have numerous benefits, according to Bisantz.

Skilled designers of products and systems, she said, continually revise their designs based on feedback they receive from tests conducted with potential users.

"There are systematic ways to do these tests," she said. "To develop a good voting system, for example, you would test it with a good cross-section of voters in the country."

She explained that a basic human-factors principle is that a good design should let the person who is interacting with the product or system identify and correct an error he or she has made before it becomes critical.

"For example, a computerized system could ask a voter who had not cast a ballot for a specific office if they had really intended to leave that blank. Maybe it would ask, "You have not cast a vote for president; did you intend to vote for a president?"

In addition to cognitive issues, she added, with some voters there also may be physical challenges, because both paper ballots and voting machines assume a certain level of manual dexterity.

"You need to test the users," she stressed. "The issue of whether or not a ballot or voting machine is easy to use is not a matter of opinion. This is something about which data can be collected. It can be empirically tested and measured."

Media Contact Information

Ellen Goldbaum
News Content Manager
Medicine
Tel: 716-645-4605
goldbaum@buffalo.edu