Statistics 101

The latest crime rankings making the rounds ("City Crime Rankings: Crime in Metropolitan America", published by CQ Press), that ranks Cincinnati as the 16th most dangerous city in the nation, is so statistically unsound, they might as well have placed all the cities in a hat and drawn names to get the rankings. The AP article already mentions two flaws, and I'll add another one:

  1. Different states define different crimes in different ways. So right there your data point is inconsistent.
  2. The study assigned a crime rate score to each city with zero representing the national average, but the study excluded Chicago, Minneapolis, and other Illinois and Minnesota cities. What kind of average is that? That's like finding the average income in America and leaving Bill Gates out (which by the way is why median is usually more useful).
  3. The FBI crime data the study uses is based on per-capita crime. But in metro cities that have annexed surrounding suburbs such as Indianapolis, Columbus, and Louisville, the per-capita crime rate will always be lower than cities like Cincinnati.
I couldn't agree more with the FBI response on the findings: "Consequently, they lead to simplistic and/or incomplete analyses that often create misleading perceptions adversely affecting communities and their residents."

Cincinnati takes another hit.

No comments: