Subjects received monetary payments for each good estimate. Possible rewards were 4, 2, or 1 points if their estimates fell into the 10%, 20%, or 40% intervals

Subjects received monetary payments for each good estimate. Possible rewards were 4, 2, or 1 points if their estimates fell into the 10%, 20%, or 40% intervals around the truth; otherwise, they received no points. The rewards applied to all rounds to make sure that individuals took all of their decisions seriously. The correct answer and the rewards for all five estimates were only disclosed to the subjects after the fifth estimate. This reward structure in- troduced incentives only to find the truth and avoided incentives to conform with others. Furthermore, there were no incentives for strategic considerations. For example, there was no benefit of being better than others or of misleading others, because this did not affect individuals ’ payments. There was also no possibility to help others by deviating from the strategy to find the best estimate. Therefore, our experimental design put subjects into a situation in which they would try to get as close to the truth as possible by using their own knowledge and the estimates of others. http://www.pnas.org/content/early/2011/05/10/1008636108.full.pdf http://www.pnas.org/content/early/2011/05/10/1008636108.full...