Using Prediction markets in scientific research
Prediction markets are now applied to more than just financial markets. The classic example is using markets to predict the outcome of political races, but markets have now been applied to predicting the predictability of scientific studies. The article linked below describes the use of a prediction market to predict the reproducibility of specifically psychological studies. The participants were given an amount to invest in the market based on if they believed the results of the study to be reproducible. The results when such a market is closed can be used to more efficiently allocate funding for retesting scientific results. Because the market is only open to experts in the field, the collective knowledge of this group is likely to predict the legitimacy of the studies they invest in. The accuracy of this market was around 70% for predicting the reproducibility of 44 psychological studies.
This studying indicates that prediction markets have significant untapped potential. When the pool of buyers in the market is made up of people knowledgeable in a given field, they are inclined to use their information to buy the must undervalued “shares,” thus driving the price back up.This study applies this methodology to specifically psychological studies, so I’d be interested to see what other scientific fields prediction markets would be able to accurately gauge. For instance, applying this method to particle physics publications seems less likely to succeed as the results of such experiments are much more deterministic than that of a psychological study.
http://phys.org/news/2015-11-results-good-true.html