How do people utilize information from outside sources in their decisions? Participants observed a signal-plus-noise or noise-alone event and then made a yes-no decision about whether a signal had occurred. Participants were provided with two information sources to aid decision making. Each source consisted of four components that provided estimates of signal likelihood. In Experiment 1, the two sources had equal overall accuracy but differed in the expertise and internal correlation of their components. A regression analysis indicated that participants overweighed the high-expertise-high-correlation source. This bias occurred on trials when the aggregate opinions of the sources disagreed. In Experiment 2, both the overall accuracy of the source and its components were manipulated. Participants overweighed information from the higher accuracy source. These biases reflect people's sensitivity to across-trial and within-trial differences in the accuracy and internal consistency of information sources. Experiment 3 provided additional evidence supporting these conclusions.