Stats Made Easy

Practical Tools for Effective Experimentation

Monday, December 11, 2006

Murderous statistics?

An assistant criminology professor at St. Cloud State University claims that a string of drowning deaths of white, male students from Upper Midwest colleges does not exhibit a random pattern based on location, race and other characteristics, including the phase of the moon. However, an article by Todd Richmond of the Associated Press cites opposing views by two professors from University of Wisconsin-La Crosse who said that "It is often harder to accept explanations that hit close to home — explanations that involve actions we ourselves have engaged in that put us at risk." Nevertheless, it is tempting to speculate, as a criminology student at St. Cloud does, that a predator prowls the Interstate 94 highway making stops at near-campus bars to look for inebriated young men.

Geographic profiling evidently is well-accepted as an enforcement tool as evidenced by its use by the Royal Canadian Mounted Police (RCMP). A Canadian police detective, Kim Rossmo, wrote the book on this subject. He details how analyis of the location and distribution of crimes can help pinpoint a murderer such as the Yorkshire Ripper. According to Rossmo, there is no such thing as a "random homicide." The problem in this case though is whether the deaths are due to homicide or random accidents to a susceptible population -- young male college students in river towns.


Post a Comment

<< Home