Lies, Damn Lies, And Statistics

The rise of empirical methods is one of the most marked recent trends in the legal academy.  Data-driven scholarship is transforming fields as diverse as intellectual property, criminal law, and environmental law.  Professor Dan Kahan, of Yale Law School, is a leader in the empirical study of law.  In an October 23, 2013, article, entitled "Statistical Fluke? Researcher's Observations on Tea Party and Science Spark Political Frenzy," Science Magazine discussed recent research by Kahan discussing correlations between political beliefs and scientific knowledge.  As the article points out,

It started innocently on 15 October. On his blog, Kahan posted an informal analysis of survey data that compares people’s comprehension of scientific concepts and their political outlook. The data were gathered from a large U.S. study of how people perceive the risk of vaccination. And when Kahan crunched the numbers, they revealed a small correlation between science comprehension and political leaning. One finding: Those who identified themselves as “liberal” tended to have greater scientific comprehension than those who self-identified as “conservative.”

This finding set the innumerate cat among the credulous pigeons.  One of the article's central insights is that statistical analysis continues to defy competent interpretation by many people, including those who should know better.  More than a century after Mark Twain attributed the phrase to Benjamin Disraeli, widespread incomprehension of statistical analysis lends credence to the observation that "There are three kinds of lies: lies, damned lies, and statistics."