News Release

Butterfly ballots confusing and biasing say researchers

Peer-Reviewed Publication

Penn State

University Park, PA – Butterfly ballots like those used in Palm Beach County, Fla., for the 2000 presidential election are confusing and biasing, and election officials would do well to consider social science theories and methodology to help prevent future controversy, according to a team of social psychologists from the University of Alberta, Canada, and Penn State.

"It is unclear whether a biasing ballot format does or should have legal standing in adjudicating disputes after an election," the researchers write in the Dec. 7 issue of the journal Nature. "Given the centrality of elections to the democratic process, it is remarkable that biasing formats continue to be used."

The researchers asked if the butterfly ballot used in Palm Beach County could have led to confusion and caused people who intended to vote for Gore to vote for Buchanan. The team includes Robert C. Sinclair, associate professor of psychology; Sean E. Moore, Carrie A. Lavis and Alexander S. Soldat, graduate students, department of psychology, University of Alberta; and Melvin M. Mark, professor of social psychology and senior scientist, Institute for Policy Research and Evaluation at Penn State. To test the question, the researchers prepared butterfly and single-column ballots with the names of leaders of ten Canadian political parties and a space for write-in candidates. The butterfly ballot placed the leaders of the two dominant parties in Edmonton in the first two places on the left-hand column. These are the places that Bush and Gore held respectively on the Palm Beach ballot. The leader of a party expected to receive few votes in the upcoming Canadian election was in the first position on the right-hand column, the position occupied by Buchanan on the Palm Beach ballot.

The researchers then asked Canadian college students to vote using the ballot. The 324 students randomly received either a butterfly or single-column ballot and voted by filling in the appropriate space in pencil. They then asked the students to assess the confusion level of their ballot and to write down which candidate they intended to voted for. The 161 students who used the butterfly ballot rated the ballot more confusing than the 163 students who used the single-column ballot. However, none of the students using either ballot voted for the wrong person.

"Although greater confusion might be expected to lead to higher error rates, we were not surprised by the lack of error in the sample because it involved students skilled at completing complex scoring sheets," the researchers said.

The researchers then took the ballots that were identical in structure to the Palm Beach ballots to a Canadian shopping mall where 116 individuals ranging from 19 to 86 years old marked ballots in a mock voting station. They also asked the participants to rate the confusion of the ballots and to write the name of the person they intended to vote for. Again, those using the butterfly ballots found them significantly more confusing than those using the single-column ballots.

While none of the 59 people using the single-column ballots made an error in voting, 4 of the 53 people using butterfly ballots voted for a different candidate than they intended. Three of the erroneous votes were intended for the candidate in the Gore position.

"The results from these two groups indicate that the butterfly ballot is more confusing than a single column ballot," the researchers say. "Our study on the second group of voters suggests that the butterfly ballot may cause systematic errors in voting which could cast doubt on the validity of the results from the Palm Beach County vote."

The researchers suggest that low-cost application of social science theory and methods would help to prevent such controversy in the future.

###

Editors: Dr. Mark may be reached at 814-863-1755 or at by e-mail m5m@psu.edu.

Public Information: A'ndrea Messer, 814-865-9481 or 814-867-1774 (h) aem1@psu.edu


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.