Voters using smartphones made fewer errors in mock election

Voters who cast their ballots via smartphones made fewer errors than they did when voting via traditional methods in a mock election, according to new research from psychologists at Rice University.

MICHAEL BYRNE

In a first-of-its-kind study, co-author Rice Professor of Psychology Michael Byrne examined how smartphone-based voting systems can be incorporated into the current large-scale voting process. The study, “Toward More Usable Electronic Voting: Testing the Usability of a Smartphone Voting System,” found that while there are no consistent differences in efficiency and perceived usability between the smartphone-based system and other voting systems, smartphone owners made fewer errors on the mobile voting system than when they used traditional voting methods.

According to Byrne, many U.S. counties have incorporated electronic voting technology, largely in response to well-publicized challenges related to older mechanical and punch-card models. He said that although these updated systems have solved some usability problems, they present a new set of issues for voters unfamiliar with the technology.

“Current electronic voting systems have numerous issues — from usability and accessibility to security to the fact that many of them are nearing the end of their life cycle — and there are few good certified alternatives currently on the market,” he said.

PHIL KORTUM

For the study, the researchers designed a mobile voting system optimized for use on a smartphone and tested its usability against traditional voting platforms. They asked 84 participants — 48 of whom reported owning a smartphone at the time of the experiment — to engage in a series of mock elections using different voting methods. The subjects ranged in age between 18 and 68 years and had varied voting histories and educational backgrounds. Ninety-two percent of participants had voted in fewer than six national elections, and 88 percent of participants had voted in fewer than six non-national elections. Seventy-five percent of the participants had either an associate’s or a bachelor’s degree.

The researchers found that the 48 smartphone owners made fewer errors on the mobile voting system than they did via the traditional system. There were no significant differences in error rates among nonsmartphone users.

Phil Kortum, assistant professor of psychology at Rice and a study co-author, said there are numerous potential benefits to using a smartphone-based voting system.

“Nobody likes to wait in line at the polling place, and so mobile voting offers the opportunity to cast votes when and where it is convenient for the voter,” Kortum said.

Kortum noted that despite these benefits, implementing smartphone voting as an anytime, anywhere system will require addressing substantial security and authentication problems that may not be solvable — and certainly not in the near future.

“Ongoing research is needed to develop systems that allow voters to securely and anonymously submit their ballots,” Kortum said. “Creating voting systems that retain the convenience of mobile phones while still ensuring the security and anonymity we enjoy with current voting technologies will be the biggest design challenge.”

The study appeared in the January 2014 edition of Human Factors: The Journal of the Human Factors and Ergonomics Society and was co-authored by Rice graduates Bryan Campbell ’11 and Chad Tossell ’12. The research was funded by Rice University.