Human Factors and Voting Technology

Abstract

The success of any election ultimately depends on the voters' ability to accurately cast a vote using the given technology. If the user interface is confusing, the system may fail to record the intent of the voter, and essentially disenfranchise citizens in the process. In the extreme case of the fiasco surrounding the presidential election of 2000, it became exceedingly clear that the study of human factors for voting design was a necessary, yet neglected, area of research. This study presents an analysis of current electronic voting systems with regard to machine type (touchscreen versus touchscreen + keypad), ballot design (full ballot interface versus one office per page interface), and effects of age. Fifty participants ages 18–26 (M = 20.52) and fifty participants ages 64–77 (M = 68.27) were recruited from the community and asked to complete a hypothetical ballot given a list of candidates to choose, in each of four experimental conditions. Results revealed lowest error rates for both groups using the touchscreen system, and lowest errors for both groups using the one office per page interface. Touchscreen systems also revealed faster completion times than touchscreen + keypad designs for both age groups, with fastest completion times occurring in the full ballot interface. Implications are discussed.