NCJ Number
164668
Journal
Journal of Quantitative Criminology Volume: 12 Issue: 4 Dated: (December 1996) Pages: 439-444
Date Published
1996
Length
6 pages
Annotation
While this study provided some evidence that dramatic changes in temperature were related to homicides, there was no association between the number of homicides and maximum daily temperature in a Baltimore, Maryland, sample of 1,462 days.
Abstract
Less than 0.5 percent of the sampled days had maximum temperatures colder than 20 degrees F and only 1 percent had maximum temperatures of 24 degrees F or colder. Multiple homicide days were also unusual, with only about 7 percent of days having two or more homicides. Using the sample marginal distributions and assuming the number of homicides is independent of maximum temperature, the authors conclude that the probability of a homicide occurring below hypothetical cold floor threshold values is very small in Baltimore. Specifically, the probability that a randomly selected day in Baltimore will have both one or more homicides and a maximum temperature at or below 19 degrees F is about 1 in 1,000 days. The failure to find evidence of a cold floor effect in Baltimore does not necessarily imply that no cold floor exists in human populations. Inferences are limited by local circumstances in Baltimore, including its relatively temperate climate. Additional investigations should include cities that experience a larger number of extremely cold days. 4 references, 3 tables, and 1 figure