Ambient temperature and violent crime: tests of the linear and curvilinear hypotheses.

Laboratory research on the effects of temperature has led theorists to propose a curvilinear model relating negative affect and aggression. Two alternative explanations of these lab findings are proposed--one artifactual, one based on attributions for arousal. Both alternatives predict a linear relationship between temperature and aggression in real-world settings, whereas the negative affect curvilinear model predicts a specific curvilinear effect. Two studies are reported that investigated the relationship between temperature and violent crime. Both studies yielded significant linear relationships and failed to demonstrate the specified curvilinear relationship. Also, both studies yielded significant day-of-the-week effects. Implications of these findings for the study of aggression are discussed.