Source: Robert Sorkin, (352) 392-0601, sorkin@psych.ufl.edu

GAINESVILLE, Fla. --- A committee has just arrived at a final list of five applicants for an important job opening. In order to be a finalist, a candidate must receive a ‘yes' vote from a specified majority of the committee.

The committee can be sure they have chosen the best candidates if they employ a simple majority vote in their selection process.

Psychologists at the University of Florida and Indiana University have found that if accurate committee votes are needed, groups are better off using a simple majority rule, or requiring a vote of 50 percent plus one vote. The stricter, or larger, the majority needed to select a candidate, the less likely that the group will make accurate selections. The research appears in the November issue of Psychological Science.

The goal of the research was to find out if group decisions are more accurate with a simple or strict majority rule. The research will help groups adopt a majority rule that best fits their needs.

Robert Sorkin, a UF psychology professor who did the study, said, "The analysis indicated that the best decisions are made when a group uses a simple majority rule. Less accurate decisions are made with two-thirds and three-fourths rules."

Sorkin said the least accurate decisions come when a unanimous rule is enforced.

Ryan West, a researcher and psychology graduate student at UF, said this research is important because it is relevant to everyday situations.

"As a whole, we vote in elections, to pass laws and sometimes for what you want on the pizza," West said. "What kind of strategies lead to the best outcome for the group?"

The research was conducted in two ways, a mathematical analysis to predict how groups would react and a computer experiment involving groups of five and seven UF students who were paid based on the accuracy of their responses.

In their tests, researchers asked the students, as individuals and in groups, to view a series of bar graphs on a computer screen and then indicate via a yes or no vote if they thought the grouping of graphs had a high or low average height. The groups performed the experiment under a simple majority rule and a strict majority rule that required yes votes from four, five, six or seven members.

Sorkin said the computer recorded the group's responses and calculated the percentage of correct answers. The results showed that answer accuracy was highest in the groups when a simple majority rule was used to determine a group vote. This result agreed with the mathematical model.

"There were one or two voters who realized that, under stricter conditions, the group was more likely to not reach the needed majority which was a cause of the decline in the group's performance," West said.

Sorkin said with stricter conditions, the performance accuracy can drop by as much as 10 percent.

The research addressed only groups whose members have limited interaction with each other before voting takes place. There is no discussion or preliminary ballot.

West said allowing the groups to talk and interact introduces new information and new variables, such as persuasion and leadership, that are hard to account for in experiments.

For groups whose members have little interaction, Sorkin recommends using a simple majority for voting purposes.

"Obviously, some groups want their decisions to be conservative," Sorkin said. "Rather than impose a strict majority rule, it would be better to convince their members to be more conservative about their individual votes."

Note: If no author is given, the source is cited instead.

• Racial Issues
• Relationships
• Spirituality

• Illusion of control
• Psychometrics
• Experimental economics
• List of cognitive biases
• Confirmation bias
• Social psychology