The Usability Implications of Long Ballot Content for Paper, Electronic, and Mobile Voting Systems
Byrne, Michael D.
Doctor of Philosophy
In the 2008 United States presidential election over 131 million ballots were cast. A substantial fraction of those ballots, approximately 23 million (17.5%), were cast as absentee ballots either domestically or by overseas and military citizens (EAC, 2008). These numbers demonstrate that a demand exists in the United States for less centralized voting procedures. One potential solution, allowing voters to cast ballots on Internet-enabled mobile devices, could potentially increase voter participation, reduce election administration costs, increase election flexibility, and provide the ability for voters to interact with familiar technology. Two experiments were conducted to examine the efficacy of a custom-designed mobile voting system as compared to more traditional voting technologies such as direct recording electronic and paper ballot voting systems. The results from experiment one suggest that displaying long ballot content as a single scrollable list may have distinct negative consequences on the effectiveness of electronic voting systems. Further, experiment one showed that candidates appearing below the fold, or not immediately visible without additional action from the voter are at a higher risk of being mistakenly voted against. The results from experiment two are largely consistent with experiment one in that they showed that a scrollable review screen led to more voting errors and that those candidates below the fold were at a distinct disadvantage.