When it comes to making a decision, or solving a puzzle, two heads really may be better than one.
A study has found the wisdom of crowds does exist, with a group of five people better able to answer questions than an individual.
The theory was tested with questions such as the height of the Eiffel tower and the number of emperors in the Roman Empire.
As pub quiz teams may suspect, groups reduced their errors by almost 50 per cent compared to one person answering alone.
Some experts have suggested that it is better not to consult with others, for fear of ‘herd think’, which has been used to explain the financial crash as well as dangerous misguided cults.
However the latest research, led by University College London, suggests people who work together share arguments and reason together to come up with the right answer.
Lead author Dr Joaquin Nvajas, now at Universidad Torcuato Di Tella in Argentina, said: ‘Many previous studies on the wisdom of crowds found that discussing with others could hinder collective wisdom because social influence leads to imitation and the power of the crowd vanishes.
‘But, to our surprise, we found that social influence and within-group deliberation actually increased the wisdom of crowds. By averaging collective estimates coming from different groups, we found that the crowd became wiser. The error decreased in approximately 50 per cent.
‘This effect was so big that just by averaging four group choices, we could outperform the wisdom of a crowd of 5,180 people.’
The study, published in the journal Nature Human Behaviour, suggests that group discussions allow people to share arguments, which increases their understanding of a problem.
It is based on 5,180 people asked general knowledge questions, which also included the number of goals scored in the 2010 FIFA World Cup and the number of lifts in the Empire State Building.
People answered alone, given 20 seconds to do so. Then they were organised into groups of five and given a minute to come up with an answer to four of the eight original questions.
The study found people reduced their errors by 49.2 per cent in a group. For example, if a correct answer was 100, and individuals put it at 200, then those working in groups would answer with a more accurate figure of 150.
When asked how they came to their decisions, people in the groups most agreed with the statement: ‘We shared arguments and reasoned together.’
The authors state: ‘Our results are in contrast to an extensive literature on herding and dysfunctional group behaviour, which exhorts us to remain as independent as possible.
‘Instead, our findings are consistent with research in collaborative learning showing that ‘think–pair– share’ strategies and peer discussion can increase the understanding of conceptual problems.’
The same authors have also found groups with individuals who hold completely opposite opinions can reach a consensus in a third of cases.
In a blog post, they wrote: ‘Because polarisation seems to have fractured our society into groups that can’t talk to each other, we found this result to be a seed of optimism.’
