New Voting Study Addresses Usability Concerns

In the debates about the use of electronic voting machines there is a refrain that direct electronic recorder (DRE) voting machines are much easier to use and much more accurate in capturing voter intent than other voting systems. A new comprehensive usability study of five commercial e-voting machines (published by the Brookings Institution) finds that we still have a long way to go in improving machine usability and that ballot design remains a challenging issue. We haven’t read the full study (Available from Brookings Press for $19.95), but here are some highlights from the news release and a story on MIT’s Technology Review.

While the study found voter confidence and satisfaction with the systems was good, error rate increased as voter tasks became more complex:

From the MIT Technology Review story, “… Bederson says that even for the simplest task–voting in one presidential race on a single screen–participants had an error rate of around 3 percent. When the task became more complicated, such as when voters were asked to change their selection from one candidate to another, the error rate increased to between 7 and 15 percent, depending on the system. Bederson notes that, although the error rate that occurred in the study may not necessarily mean that there is the same error rate in terms of actual votes on actual machines, the study does raise concern, considering how close some recent elections have been.”

The study has raised concerns within the computing community because it seemingly downplays security issues. The press release on the study clearly plays up that there is some tradeoff between usability and voter-verification. I suspect this is more likely to do with trying to generate a controversial enough angle for the press than the authors belief that completely software dependent voting systems are worry free. In general, we think this “tradeoff” view is flawed. There are clearly major security and reliability issues that justify having software-independent verification systems, just as there are clearly usability issues that need attention. Attention doesn’t need to be mutually exclusive.

The key to improving voting technology is careful engineering, research into voting technology and voter behavior, and strong standards and testing. We are a long way from meet these objectives. More attenion on all these issues should be the goal for election officials, vendors, technologists, and policymakers.