Why do we sometimes decide to take risks and other times choose to play it safe? The work is described in the March 21 online early edition of the Proceedings of the National Academy of Sciences. In the study led by John O'Doherty, professor of psychology and director of the Caltech Brain Imaging Center, 24 volunteers repeatedly participated in three types of trials: a "Self" trial, in which the participants were asked to choose between taking a guaranteed 10 or making a risky gamble with a potentially higher payoff; an "Observe" trial, in which the participants observed the risk-taking behavior of a peer (in the trial, this meant a computer algorithm trained to behave like a peer), allowing the participants to learn how often the peer takes a risk; and a "Predict" trial, in which the participants were asked to predict the risk-taking tendencies of an observed peer, earning a cash prize for a correct prediction. Notably in these trials the participants did not observe gamble outcomes, preventing them from further learning about gambles. O'Doherty and his colleagues found that the participants were much more likely to make the gamble for more money in the "Self" trial when they had previously observed a risk-taking peer in the "Observe" trial.