Two Tales of Randomness

People encounter randomness everywhere in life, but they often fail to adequately account for it when they process quantitative information. Two projects show that people typically underappreciate the impact of randomness, leading to a very naïve attribution process. A first project finds that, due to a failure to understand regression to the mean, people assume that the option that looks the best is the best, leading to dissatisfaction with outcomes of decisions based on unreliable information. This “maximizer’s curse” is purely statistical in nature, and is exacerbated when the number of options increases. A second project shows that people are reluctant to attribute differences to good/bad luck, especially when these differences seem larger. As a result, they think differences are more likely to replicate when they look larger, leading to more confident decisions.