Bandit problems provide an interesting and widely-used setting for the study of sequential decision-making. In their most basic form, bandit problems require people to choose repeatedly between a small number of alternatives, each of which has an unknown rate of providing reward. We investigate...