Usability In The Free World

When there is a risk that usability issues with ballot papers could decide the winner in the US Presedential Elections, surely it is time to sit up and take notice.

Next time somebody tells you usability doesn't matter, consider events in Florida over the last week. Specifically, it's worth reflecting on the fact that the next President of the United States may be decided due to the poor design of one ballot paper in Palm Beach County. While it is outside of our usual subject matter, this particularly story is just too topical to pass up. And it exhibits all the hallmarks of a classic usability blunder:

The ballot paper design was checked repeatedly, distributed to users for approval and signed off by the authorities. But it was never TESTED on the people who would be asked to make sense of it. It is one thing to casually glance at a proposed ballot in the paper, another to be forced to interact with it in a polling environment.

The ballot paper was inconsistent and counter-intuitive. Most people can agree that it doesn't look too complicated now, but for many voters the arrangement of the paper went against everything they expected to occur. More specifically, most voters would assume that the two main candidates would occupy the two top holes, and would have used such a system for most of their lives. Placing a third candidate BETWEEN Bush and Gore is asking for confusion. What is created here is a conflict between accumulated knowledge and the evidence of the ballot paper itself - the result is confusion.

Sections of the media are already engaged in the process of 'blaming the user'. The argument is a simple one - "I can understand this ballot paper - if you can't then your stupidity is your own problem." This common reaction to such a situation (and one frequently encountered when designing web or software interfaces) is a refusal to recognize that an observable problem exists. It doesn't matter why the problem occurs - it only matters that it can be solved through design more easily than by re-educating thousands of voters (or customers).

The importance of understanding the user and the environment they work in is at the heart of the controversy. Those of us blessed with 20/20 vision, looking at a ballot-paper laid flat, may well find it hard to believe it caused a problem. But elderly voters looking at the ballot at an angle, in a polling booth, may not have enjoyed such an unambiguous view.

It is difficult to say, without testing, exactly how such an occurrence can be avoided. It seems apparent that many voters will read down the first column and look instinctively for the second box, perhaps not even noticing the arrows. This probably indicates that any form of 'butterfly' layout is fundamentally flawed. There are a myriad of other potential reasons for voter confusion - many already suggested elsewhere. Either way, the lesson must be that ballots should be tested rigorously before being used for real.

Although we have no stated preference regarding who wins the election, it would be nice to see ballots designed with voters in mind.