Testing the reproducibility of social science research
A team co-led by Gideon Nave of the Wharton School replicated 21 high-profile social science studies and found discrepancies with the original research. Researchers betting in prediction markets, however, were quite accurate at predicting which findings would replicate and which would not.