Research | UW news blog
October 26, 2022
When groups of people need to make a decision, they often conduct a straw poll to test opinions before the official vote. New research from the University of Washington shows that a particular voting method has been shown to be more effective than others in determining the best choice.
In a study published Sept. 28 in Academy of Management Discoveries, researchers found that groups who used “multivoting” in informal voting were 50% more likely to identify the correct option than those who used plural or ranked voting.
Multivoting gives people multiple votes to cast across all options. The reality show American Idol uses multivoting, giving fans 10 votes each. They can use all 10 for their favorite candidate or split their votes between two or more. For this study, the students received 10 votes, which they could distribute among three choices.
Majority voting, in which voters must select one option, is most commonly used in political elections. Increasingly popular in some local and state political elections, ranked voting allows people to list their preferences from start to finish. It is also used to determine Oscar winners.
Michael Johnson, co-author and professor of management at the UW Foster School of Business, said that most achievement groups, who want to be sure they are making the best decision, do multivoting. The researchers don’t think it would work for political elections, mainly because it would be too tiring to spread votes across a multitude of options.
“We see multivoting as primarily useful for decision-making groups in the workplace,” Johnson said. “Whenever groups feel it is important to make a right decision, use multivoting as an unofficial vote, see the distribution and then discuss. It works when people are motivated to vote in line with what they really think, rather than trying to strategically vote against someone else.”
The UW study was based on the “pursuit teams” developed by the Department of Homeland Security after 9/11. The purpose was to link the results of multiple intelligence agencies to track down potential terrorist threats.
In this study, researchers asked 93 groups of college students to simulate counterterrorism support teams and identify which of the three suspects posed the greatest threat. The student groups received information on three terrorists, but no group member had complete information on a suspect. Students had to share information to correctly identify the greatest threat.
Teams were split into thirds, creating an even number of groups with ranked voting, plurality, and multivoting. All groups took a preliminary, unofficial vote to see members’ first thoughts on the terror suspects. After the unofficial vote, they looked at the results and discussed the suspects. If the students combined the information well, they would be able to identify a terrorist who clearly posed a greater threat. The teams then gave their final verdict.
Only 31% of the plural teams chose the most threatening suspect in the final vote, about as much as if it were left to chance. In the unofficial vote, 6% of teams had a majority of members identifying the right suspect. This is less than the 11% expected by chance.
The ranking choice voting did not fare much better. In the final vote, 32% of the teams identified the right suspect. In the unofficial vote, 7% of the groups had a majority of members ranking the right suspect as the most threatening.
“We were surprised that the ranked groups didn’t outperform the plural groups,” Johnson said. “There is a lot of evidence today, particularly in politics, that rank voting produces outcomes that better match voter preferences than majority voting. That’s why we’ve seen so many political elections move toward ranked voting.
“But ranking is generally better at uncovering people’s true preferences and not necessarily finding the exact right answer. When people make decisions at work, you’re more concerned with getting it right than making sure it reflects everyone’s opinion.”
The multivoting groups started out stronger, with most members voting for the most threatening suspect in 30% of the groups. In the final vote, 45% of the teams identified the most threatening suspect.
The researchers found no evidence that the discussions in the multivoting groups differed in any way from the other two voting conditions. Instead, the benefit of multivoting occurred before each discussion, as students processed the information more thoroughly and viewed intelligence more critically.
“The real discovery, and what we didn’t expect, was that multivoting groups were more specific before discussing,” Johnson said. “We just assumed they were all kind of the same before the discussion and would improve by the end. If people have the opportunity to say, ‘I kinda like option A, but I also like option B’, that might make them think more before discussing what would help them make the right decision.”
Other co-authors were Eli Awtrey from the University of Cincinnati and Wei Jee Ong from the National University of Singapore.
For more information, contact Johnson at [email protected]
Keyword(s): Foster School of Business • Michael Johnson