In the first study, 751 subjects answered questions regarding their crime judgments as controls, stereotypic perceptions of muggings, and recall of violent crimes. The findings showed that the subjects believed stories of minor injuries sustained from a crime that were told by interpersonal sources more than they believed media stories involving major crime-related injuries. In addition, people who exchanged information about crime with interpersonal sources acquired a more realistic portrait of the offense. When asked for their opinions on judicial sentencing, the subjects tended to rely on crime stereotypes, as portrayed in the media. The second study provided strong evidence that people use an availability heuristic in forming punishment preferences to abstract cases, often resulting in opinions based on a biased sample of cases. However, the recall advantage for severe harm is eliminated when people are presented with stories about minor harm that are highly concrete or contextually distinct. 4 tables, 7 notes, and 43 references