Citation

Cohn EG, Rotton J, Peterson AG, Tarr DB. J. Appl. Soc. Psychol. 2004; 34(8): 1652-1674.

Copyright

(Copyright © 2004, John Wiley and Sons)

DOI

10.1111/j.1559-1816.2004.tb02792.x

PMID

unavailable

Abstract

A review of time-period research on temperature and aggression led the authors to hypothesize that southern subculture and sociodemographic variables are responsible for geographical differences in 11 southern homicide rates. It was also hypothesized that temperature's correlation with nonhomicidal violence would be stronger in small than large cities. These hypotheses were tested by obtaining cross-sectional data on 10 temperature indexes, 11 sociodemographic controls, percentage southern horn, and crime rates in more than 300 U.S. cities in 1990. Partial correlation analyses indicated that percentage southern born and sociodernographic variables rather than temperature predicted homicide rates. Moderator variable regression analyses indicated that relationships between temperature and nonhomicidal violence were stronger in small than large cities. The results are interpreted in terms of a social escape-avoidance model of criminal behavior, which predicts that low as well as high temperatures lead co-offenders to avoid social contact.