Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
People in our experiment choose between two information sources with opposing biases in order to inform their guesses about a binary state. By varying the nature of the bias, we vary whether it is optimal to consult information sources biased towards or against prior beliefs. Even in our deliberately-abstract setting, there is strong evidence of confirmation-seeking and to a lesser extent contradiction-seeking heuristics leading people to choose information sources biased towards or against their priors. Analysis of post-experiment survey questions suggests that subjects follow these rules due to fundamental errors in reasoning about the relative informativeness of biased information sources.